Iron Oxide Copper-Gold Deposits

Workshop & Core - Adelaide .......... Monday 21 November, 2005.

A one day workshop was held in Adelaide, South Australia to provide background information and a context to the deposits to be visited over the following 2 days.

The workshop covered the regional to local scale tectonic, geologic and metallogenic setting and alteration patterns associated with IOCG mineralisation in the Gawler Craton and Curnamona Province, including the characteristics of deposits not to be visited. Expert presenters included Sue Daly and Alan Maugr from PIRSA Minerals (the South Australian State Geological Survey), and Dr Roger Skirrow of GeoScience Australia (the Australian Commonwealth Geological Survey) and representatives of Teck Cominco.

These presentations were followed by a drill core inspection session at the PIRSA Drill Core Storage Facility with core from six significant IOCG occurrences/deposits in the Mount Woods Inlier, the Stuart Shelf and the Moonta-Wallaroo region, all in Gawler Craton and from the adjacent Curnamona Province. All were from occurrences that were not visited. In addition, the recent discovery hole which contains a major IOCG intersection at the Carrapateena Prospect (south of Olympic Dam) was studied.

Return to top

Our International
Study Tour Series
The last tour was

Our Global Perspective
Series books include:

Super Porphyry Cu and Au

IOCG Deposits - 70 papers

All available as eBOOKS
Remaining HARD COPIES on
sale. No hard copy book more than AUD $44.00 (incl. GST)

Olympic Dam ...................... Tuesday 22 November, 2005.

The Olympic Dam copper-gold-silver-uranium-REE ore deposit is located ~560 km NNW of Adelaide and 275 km NNW of Port Augusta, in northern South Australia (#Location: 30° 26' 24"S, 136° 53' 22"E).

Go to: Regional Setting, Geology, Structure, Alteration, Mineralisation, Geophysical Expression, Evolution and Discussion and Reserves and Resources.

DISCOVERY and HISTORY

The Olympic Dam deposit was discovered in July 1975 by Western Mining Corporation Ltd (WMC; Reeve, 1990) and was the first deposit of its type recognised, leading to the creation of the iron oxide copper-gold (IOCG) deposit model by Hitzman, Oreskes and Einaudi (1992). The discovery was the result of exploration for sediment hosted copper mineralisation, based on a conceptual model advanced by Haynes in 1972 (Haynes, 2006). This model envisaged alteration of continental tholeiitic basalt to an assemblage of albite-hematite-phyllosilicates-epidote-carbonates by deeply circulating oxidised fluids. It was postulated that this alteration process would have released copper, mainly hosted in magnetite within the volcanic rocks, to be taken into solution by these fluids. It was suggested these copper-pregnant fluids then migrated up steeply dipping faults until encountering reduced sedimentary rocks where copper was precipitated in a sedimentary trap, such as a domal structure (Haynes, 1972).

The ~1590 Ma Roopena Volcanics in South Australia were identified as a potential example of copper-depleted altered continental tholeiitic basalts. These basalts underlie the flat-lying Meso- to Neoproterozoic to Lower Palaeozoic sedimentary sequence of the Stuart Shelf. Prospective targets were delineated by geophysical data in conjunction with lineament analysis (Rutter and Esdale, 1985; O'Driscoll, 1985; Esdale et al., 2003). Coincident gravity and magnetic anomalies were considered to reflect the geophysical expression of mafic volcanic rocks in the core of a domal structure, or in a fault bounded basement high, that might satisfy the proposed model. The first drill hole at Olympic Dam was sited on such a geophysical target, developed at the intersection of WNW and NNW lineaments. The target was was developed from geophysical data from pre-competitive surveys by the Commonwealth Bureau of Mineral Resources and the South Australian Department of Mines and Energy.

The discovery diamond hole, RD1, drilled over this anomaly in July 1975, intersected 38 m @ 1.05% Cu, 70 ppm U3O8, 0.27 g/t Au, 3.5 g/t Ag at a depth of 353 m beneath the surface (Roberts and Hudson, 1983). There was an apocryphal rumour circulating in the early 1980s that the mineralisation was not visually obvious, and the interval was only assayed after the core was left in the open for some months and the fine sulphides had begun to oxidise. The next eight drill holes contained very low to low-grade copper. High-grade mineralisation was not intersected again until 15 months after RD1, in drill hole RD10, which intersected 170 m @ 2.09% Cu, 630 kg/t U3O8, 0.65 g/t Au, 3.1 g/t Ag). RD3, 4, 6 and 7 had been barren, RD5 and 8 intersected additional mineralisation containing ~1% Cu, while RD9 was lower grade. The RD10 intersection was the first indication of potentially economic grades of both copper and uranium, and provided the impetus for the initiation of a comprehensive surface drilling program that resulted in an estimate of >2 Gt @ 1.6% Cu, 0.06% U3O8 and 0.6 g/t Au. Drilling over an area of 25 km² resulted in a success rate of >80% mineralised drill holes (Roberts and Hudson, 1983). That drill program was undertaken under a joint venture signed in mid-1979 between WMC (51 %) and the British Petroleum Group (49%).

The host sequence was interpreted to be basement to the Neoproterozoic and younger sedimentary rocks of the Stuart Shelf, and to predominantly comprise a >1 km thick succession of sedimentary breccias that ranged from matrix-poor with granite clasts, to matrix-rich polymict breccias containing clasts of a variety of rock types. These sedimentary breccias were interpreted to have been deposited within a NW-SE elongated graben. The sequence was divided into, i). the Olympic Dam Formation, with five members, three of which are matrix rich; and the overlying ii). Greenfield Formation, with three members, the lower two being very hematite rich while the upper has a significant volcanic component. Pervasive hematite, chlorite and sericite alteration of varying intensity was interpreted to have affected all the members of both formations. The basement below the Olympic Dam Formation was assumed to be a granite, equivalent to that found outside of the known area of mineralisation (Roberts and Hudson, 1983).

Mineralisation was interpreted to comprise stratabound bornitic (chalcocite)-chalcopyrite-pyrite in the hematite matrix-rich members of the Olympic Dam Formation, with sulphides mainly occurring as uniform disseminations in the breccia matrix, although up to 5 cm clasts of massive sulphides were also observed. A vertical zonation of sulphides was noted, ranging from a sulphur-rich, copper-poor pyrite (-chalcopyrite) at the base of the Olympic Dam Formation, to a copper-rich and sulphur-poor assemblage of bornite (-chalcocite) at the top. No evidence of supergene enrichment or lateral zoning of sulphides was noted. However, a distinct siderite-rich zone was seen to follow the northeastern margin of the deposit, in which pyrite is the dominant sulphide. Intense chlorite alteration was interpreted to have developed at the base of the stratabound mineralisation, whilst transgressive chalcocite-bornite was observed to overprinted the stratabound mineralisation forming a 6 by up to 0.7 km NW-trending linear corridor in the centre of the proposed graben. Massive sulphide clasts, veins, clast rims and cavity infill were observed, complexly intergrown with the chalcocite-bornite. Sericite-quartz or chlorite alteration halos were observed, developed around sulphide veins. In both the stratabound and transgressive mineralisation, hematite, quartz, sericite and fluorite were observed to be the principal gangue minerals, with lesser chlorite, siderite, barite and rutile. Very fine grained uraninite, with lesser coffinite and brannerite were noted to occur in a variety of forms, whilst the rare earth minerals bastnaesite and florencite were found in and adjacent to, sulphide mineralised zones. The interpretations of Roberts and Hudson (1983) were based on relatively sparse surface drilling alone.

An exploration shaft was sunk to a depth of 500 m in 1982, and underground development through the mineralised zones commenced. Detailed underground drilling and geological evaluation followed, during which it was realised that the Olympic Dam mineralisation was not hosted by a sedimentary breccia, but was associated with intense, texturally destructive hematite and sericite alteration and brecciation of the primary host rock units, including the ~1593 Ma Roxby Downs Granite, 'bedded clastic facies' rocks, and mafic-ultramafic dykes, lavas and sills. This realisation led to a reappraisal of the deposit's origin, the most accepted of which involved a maar-diatreme setting, where fluid mixing was responsible for alteration and metal precipitation in a near-surface regime, temporally and spatially associated with the extensive ~1590 Ma Gawler Range Volcanics - Hiltaba Suite event (Reeve et al., 1990; Haynes et al., 1995; Johnson and Cross, 1995). In this model, the Olympic Dam and Greenfield formations of Roberts and Hudson (1983), broadly coincided with the altered and mineralised Roxby Downs Granite and maar crater sediments respectively. A less widely accepted model, proffered by Oreskes and Einaudi (1990, 1992), involved a faulted basin and multiple fluid sources, where the hydrothermal system and brecciation was activated ~100 m.y. after the Gawler Range Volcanics - Hiltaba event, and subsequently modified by supergene processes. Following a further major ~450 km drilling program of close spaced holes, McPhie et al. (2011, 2016) showed that the volcaniclastic and minor sedimentary units previously interpreted to have been deposited in maar craters, were a series of thick, laterally continuous 'bedded clastic facies' largely deposited below wave base in a subaqueous sedimentary basin. These volcano-sedimentary rocks have been faulted and incorporated into the breccia complex, where brecciation that was initiated along faults allowed the deep circulation of fluids. Explosive injection of multiple mafic and ultramafic dykes in a wet environment, resulted in further local brecciation (McPhie et al., 2011).

In April 1993, WMC (Olympic Dam Corporation) Pty Ltd purchased the BP Group 49% share. Then, in June 2005, the BHP Billiton Group acquired WMC Resources Ltd., and the Olympic Dam operation was merged into the former, which in November 2018 became BHP Group Limited.

REGIONAL SETTING

Olympic Dam, and the other significant known IOCG mineralised systems of the Mesoarchaean to Mesoproterozoic Gawler Craton, are hosted within Palaeo- to Mesoproterozoic rocks, and are distributed along the eastern rim of the currently preserved craton to define the Olympic IOCG Province (Skirrow et al., 2007). Olympic Dam lies below the Neoproterozoic Stuart Shelf, where >300 m of flat lying, largely barren, Neoproterozoic to lower Palaeozoic sedimentary rocks unconformably overlie both the craton and the ore deposit. However, regionally, sections of this cover sequence do contain sediment hosted Cu mineralisation, some of which has been mined (e.g. Mount Gunson, Cattle Grid). Some 75 km to the east, this cover sequence expands across the major NNW trending Torrens Hinge Zone at the edge of the craton, into the thick succession of the north-south aligned Neoproterozoic Adelaide Geosyncline rift basin. This rift basin masks the mid- to late-Palaeoproterozoic suture between the Gawler craton and Palaeo- to Mesoproterozoic Curnamona Province to the east.
For a more detailed description of the tectonic and geologic setting of the Olympic IOCG Province and the Gawler Craton, see the separate Gawler Craton and Olympic IOCG Province record.
In summary, the oldest rocks in the Gawler Craton include the ~3.40 to 3.15 Ga Cooyerdoo Granite and underlying Palaeoarchaean TTG basement, followed by further felsic magmatism represented by the ~2820 Ma Coolanie Gneiss. These units form the core of the craton, but are only exposed as windows through younger sequences on the northeastern Eyre Peninsular to the west of the Olympic IOCG Province. Extensive Neoarchaean complexes deformed by early Palaeoproterozoic orogenesis comprise the gneissic Mulgathing and Sleaford complexes in the northern and central-southern sections of the craton. The southern and eastern Gawler Craton, within and to the immediate west of the Olympic IOCG Province, is underlain by the 1.96 to 1.85 Ga Hutchison Group volcano-sedimentary rocks, 1855 to 1845 Ma Donington Suite granitoids and the 1.76 to 1.74 Ga volcano-sedimentary Wallaroo Group and equivalents. The major 2.48 to 2.42 Ga Sleafordian and 1.73 to 1.69 Ga Kimban orogenies/metamorphic events affected the Archaean and Palaeoproterozoic units respectively (Hand, Reid and Jagodzinski, 2007).
The eastern Gawler Craton contains a major lithospheric boundary, interpreted to be a palaeo-suture (Hayward and Skirrow, 2010; Betts et al., 2015). The precise location of this boundary is uncertain in the Olympic Dam area. Betts et al. (2015) suggest this suture corresponds to the NW-SE Elizabeth Creek Fault that forms the western margin of the Olympic IOCG Province. However, a major linear discontinuity in magnetic and gravity data beneath the Neoproterozoic Adelaide Geosyncline intracratonic rift complex to the east is also suggested to represent this structure (Hayward and Skirrow, 2010, and sources quoted therein), an option that is more likely. Betts et al. (2015) argue that the suture initially formed at ~1850 Ma, during the Cornian Orogeny, when the Gawler Craton was accreted to the North Australian Craton, a remnant of which is the Curnamona Block. This orogeny is reflected by the emplacement of the Donington Suite intrusives which were extensively developed along the northern Olympic IOCG Province. Betts et al. (2015) further suggest that the South Australian/Gawler and North Australian Craton were subsequently rifted apart during a period of extension and basin development (including deposition of the Wallaroo Group) that lasted until ~1740 Ma, and that the two terranes were finally re-sutured during the 1730 to 1690 Ma Kimban Orogeny.
The Mesoproterozoic succession is dominated by the ~1595 to 1545 Ma Hiltaba/Gawler Range silicic large igneous province (Jagodzinski et al., 2021), which is mainly composed of the largely undeformed, subaerial, Gawler Range Volcanics (GRV; Giles, 1988; Blissett et al., 1993; Allen et al., 2003, 2008).The GRV is estimated to have totalled ~90 000 km³ of magma (McPhie et al., 2008). These volcanics, although bimodal, are largely felsic with lesser mafic varieties and were predominantly deposited between 1595 and 1575  Ma in two suites, a Lower, ~1592 Ma, and Upper, ~1588 Ma Gawler Range Volcanics. They effectively bracket the extensive A- and I-type felsic granitoids and lesser mafic rocks, including gabbros, of the 1598 to 1583 Ma Hiltaba Suite (Fanning et al., 1988; Flint, 1993), with the former dominating in the Olympic IOCG Province. The Lower GRV includes dacite to rhyolite, ignimbrites and tuffs with minor to locally significant basalt erupted from discrete volcanic centres. Basaltic to andesitic rocks are most prevalent along the western margin of the Olympic IOCG Province, from ~50 km SE of Prominent Hill to ~100 km south of Carrapateena. The upper GRV comprises three regionally extensive, 200 to 300 m thick, high temperature (900 to 1100°C) felsic lava flows. The various volcanic centres of the Lower GRV are considered to be derived from fractional crystallisation and crustal contamination of mantle melts (Fricke 2005; Creaser 1995; Stewart 1994; Giles 1988), whereas the flows of the Upper GRV are isotopically homogenous and define a single fractionation trend, suggesting that they were derived from the periodic tapping of a single stratified magma reservoir (McAvaney and Wade 2015; Stewart 1994).
Through Pb-isotopic analyses of alkali feldspar in the Hiltaba Suite and GRV, Chapman et al. (2019) revealed a craton-wide isotopic zoning. Pb-isotopic zonation is controlled by the varying interaction of a parental melt derived from i). a lithospheric mantle-like enriched reservoir which exerts a strong control on Pb-isotopic compositions around the central-western Gawler Craton (Nuyts Terrane); and two crustal end-member reservoirs which are ii). an unradiogenic reservoir which represents granulitic lower crust, and appears to be spatially limited to the northwest of the Gawler Craton; and iii). a radiogenic reservoir, which is more prevalent in the eastern Gawler Craton, coincident with the IOCG province and probably represented by monazite-rich metasediments.
The Hiltaba Suite and Gawler Range Volcanics are oldest in the immediate Olympic Dam area, with a weighted average age of 1592.62 ±0.66 Ma (Jagodzinski et al., 2021). Most of the magmatism dated has returned ages of between ~1595 and 1575 Ma, with ~14% younger, between ~1575 and 1545 Ma. The latter are concentrated around the margin of the Gawler Craton in the Mount Painter and Mount Babbage Inliers, as well as on Yorke Peninsula, Peake and Denison Ranges and Fowler Domain, and in the Curnamona Province (Jagodzinski et al., 2021).
McPhie et al. (2011) showed via analyses of melt inclusions in quartz phenocrysts from rhyolite of the Gawler Range Volcanics, that the silicic magmas of this large igneous province were unusually rich in F, containing up to 1.3 wt.% F. As a consequence they infer any associated fluid would contain abundant hydrofluoric acid, the most corrosive acid known, accelerating any fluid-rock and fluid-metal interaction.
These rocks were locally affected by the 1565 to 1540 Ma Kararan Orogeny that led to the reactivation of faults in the northern and western Gawler Craton (e.g., Ehrig et al., 2017).
Around Olympic Dam, in the eastern Gawler Craton, the rocks of the large igneous province are overlain by the continental-derived sedimentary rocks of Calymmian Pandurra Formation, deposited within the Cariewerloo Basin between 1575 and 1425 Ma associated with an inferred extensional regime that bracketed that period. In situ Pandurra Formation has not been formally recognised immediately below the Neoproterozoic unconformity at Olympic Dam. However, as detailed below, a narrow sliver of well-sorted quartz-rich sandstone containing detrital zircon grains with an age spectrum identical to that the Pandurra Formation is structurally interleaved with the host sequence within the Olympic Dam mine (Ehrig et al., 2017). It is also possible that the 'bedded clastic facies' within the Olympic Dan Breccia Complex are the precursors of the Pandurra Formation. The Pandurra and older formations were intruded by the 180 to 250 km wide, NE-SW trending, ~825 Ma, mafic-ultramafic Gairdner Dyke Swarm (Zhao and McCulloch, 1993; Zhao, McCulloch and Korsch, 1994), and unconformably overlain by the, extensive, mainly flat lying, Neoproterozoic to Cambrian sedimentary rocks of the Stuart Shelf (Preiss, 1993).

GEOLOGY

The Olympic Dam iron oxide Cu-U-Au-Ag deposit is truncated by a major unconformity at the top of the Mesoproterozoic basement, below the cover sequence described above. The delineated resource covers an area of ~6 x 3 km and is relatively thin with a vertical extent that generally extends no more than 800 m below that basement surface, except in the southeastern segment where it has been tilted and down-faulted to depths of as much as 2 km or more below the unconformity.
The iron oxide Cu-U-Au-Ag mineralisation of the deposit is hosted entirely within the Olympic Dam Breccia Complex that is principally developed within the ~1593 Ma Roxby Downs Granite of the regional Hiltaba Suite (Reeve et al., 1990). Other less abundant rock units within the Olympic Dam Breccia Complex include felsic and mafic to ultramafic volcanic rocks of the marginally older, but comagmatic Gawler Range Volcanics, as well as a thick succession of overlying 'bedded clastic facies' rocks that are marginally younger than the Roxby Downs Granite (McPhie et al., 2011, 2016). The Roxby Downs Granite intrudes the Palaeoproterozoic Wallaroo and Hutchison groups that comprise meta-sedimentary and meta-volcanic rocks and iron formations, as well as variably deformed Donington Suite megacrystic granite, and ~1590 Ma Gawler Range bimodal ultramafic to felsic lava and sills, and ignimbrites. All of these rocks are unconformably overlain by an ~350 m thick cover sequence of the Stuart Shelf that comprises flat-lying, undeformed, post-brecciation and post-mineralisation Neoproterozoic to Cambrian sedimentary units. The Neoproterozoic Gairdner Dyke Swarm dolerite and basalt intrude all basement lithologies at Olympic Dam (Reeve et al., 1990; Ehrig, McPhie and Kamenetsky, 2012; Huang et al., 2015).
Ehrig (2019) emphasised that during and over the succeeding billion years after its deposition, the Olympic Dam deposit has been substantially structurally dislocated, such that the present orebody geometry does not reflect the pre- or syn-mineralisation structural architecture, but rather reorganised and partially remobilised remnants. During this same extended period, the deposit has been influenced and modified by a number of tectonic events and pulses of hydrothermal fluid circulation, including, in particular, upgrading by addition of further uranium between 0.7 and 0.5 Ga (Ehrig et al., 2021).

Olympic Dam Geology Plan

Click here to expand -or- here for a location map

Olympic Dam Cross Sections - NW

Click here for cross section location map

Olympic Dam Cross Sections-SE

The geology within the deposit area has been subdivided as follows, after Ehrig et al. (2012, 2017) and McPhie et al. (2016):
NOTE: Sericite and muscovite are used interchangeably in this and source descriptions, and include sericite, fine-grained muscovite, illite, phengite and paragonite.

Pre-brecciation lithologies
RDG - Roxby Downs Granite, which is principally an undeformed, pink to red, medium- to coarse-grained (<2 cm), hypidiomorphic, equigranular to slightly porphyritic granite that contains ~50 wt.% alkali feldspar, 21 to 25% quartz, 17 to 23% sodic plagioclase and 5 to 7% biotite and amphibole, accompanied by accessory to trace igneous magnetite, titanite, apatite, zircon, allanite, fluorite, tourmaline, calcite-ankerite, synchysite, uranothorite and trace sulphides (Ehrig, McPhie and Kamenetsky, 2012). Alkali feldspars (~Or55 Ab43 An2) which are coarsely perthitic, and plagioclase with two compositions (~An12-20 and ~An27-34) display both rapakivi and anti-rapakivi, and occasionally pseudogranophyric textures (Kontonikas-Charos et al., 2017). Unaltered Roxby Downs Granite is characterised by the presence of biotite, K2O/Na2O ratios of 5.5:3.5, magnetic susceptibility range of ~25 to 50 x 10^-3 SI units, and a bulk dry density of ~2.65 to 2.70 g/cm³. The Roxby Downs Granite has been dated at 1593.3 ±1.2 Ma (SHRlMP, U-Pb zircon; Jagodzinski, 2014) and 1593.87 ±0.21 Ma (Cherry et al., 2018). The age and contact relationships of the Roxby Downs Granite suggest it intruded the overlying, broadly comagmatic Gawler Range Volcanics which have been dated at 1594.73 ±0.30 Ma; (Cherry et al., 2018).

Felsic Volcanic Units - Although no in situ felsic lavas have been mapped at Olympic Dam, remnants of feldspar-phyric felsic lavas occur as angular porphyritic fragments in hematite breccias and as clasts in some of the 'bedded clastic facies' (McPhie et al., 2011, 2016; Ehrig, McPhie and Kamenetsky, 2012). Zircons from these felsic fragments have been dated at ~1590 Ma (Johnson and Cross, 1995) and are interpreted to have been derived from the regionally extensive Gawler Range Volcanics (Ehrig et al., 2017).

Pre- to Syn-brecciation Units
Bedded Clastic Facies, as preserved, occur as an up to 350 m thick packages within, but mainly near the top of, the Olympic Dam Breccia Complex, immediately below the Neoproterozoic unconformity, although large blocks have been down-faulted to deeper levels. Contacts between these facies and the hematite-rich breccias are commonly faulted, but are elsewhere gradational. Determination of the relationship is complicated by the intense alteration that has overprinted the contacts. However, at gradational contacts, the 'bedded clastic facies' are commonly brecciated and separated from hematite-rich or granite-rich breccias by intervals composed of fragments of the 'bedded clastic facies' dispersed through hematite-rich or granite-rich breccia, whilst domains of hematite-rich breccia also cut across the bedded sedimentary facies (McPhie et al., 2011). Never-the-less, the 'bedded clastic facies' rocks are markedly different to the hematite-rich breccia, distinguished by their overall finer grain size and well-defined planar bedding.
McPhie et al. (2011) note that these facies are composed of laminated to very thinly bedded mudstone, thinly to medium-bedded sandstone, and thickly bedded conglomerate. They include both red, hematite-rich, and green, chlorite-rich mudstone and sandstone intervals that can be tens of metres thick. The red, hematite-rich sandstone intervals include abundant granite fragments and granite-derived feldspar and quartz, whilst those in the green, chlorite-rich intervals are mainly composed of mafic volcanic fragments, as well as containing detrital chromite. Whilst the mafic volcanic fragments match those of the Gawler Range Volcanics, the chromite grains do not, and appear to have been derived from separate sources and mixed by surface sedimentary processes prior to final deposition. Volcanic quartz (<5 modal %) is found in most of both red and green sandstone, although while such quartz-bearing volcanic units occur in the Gawler Range Volcanics, none are exposed proximal to Olympic Dam. The conglomerate beds comprise well-rounded, moderately sorted, feldspar-phyric felsic and mafic volcanic clasts and minor granite fragments (McPhie et al., 2011). These facies, which have been variably modified by hydrothermal activity, are interpreted to be mainly sourced from the Gawler Range felsic and mafic volcanics and Hiltaba Suite felsic intrusions, although Archaean and Palaeoproterozoic zircons in well-sorted quartz-rich sandstone indicate that distal older Gawler Craton basement also contributed detritus (McPhie et al., 2011).
Accumulation of the 'bedded clastic facies' was at least partly contemporaneous with explosive rhyolitic eruptions from vents that were most likely tens of kilometres or more distal, indicated by bubble-wall shards found in tuffaceous mudstone facies (McPhie et al., 2011). As such the sequence overlaps the waning stages of the Gawler Range Volcanics. The laminated mud and graded sand beds are typical of below-wave-base subaqueous settings, whilst the conglomerate beds are likely to be deposits from high-concentration gravity currents such as debris and turbidity flows. The preserved sections of these rocks suggest that the un-compacted sedimentary thickness was >500 m. The area covered by the sedimentary basin is unknown, although the dominance of subaqueous below-wave-base facies, its substantial thickness, and presence of exotic detritus imply it was much more extensive than the preserved bedded sedimentary facies at Olympic Dam, and is not consistent with deposition confined to a maar crater, either with regard to area, thickness or facies of the sedimentary sequence. The extent and character of this basin are interpreted to imply the the Olympic Dam hydrothermal system operated beneath, and partly within, the basal sequences of an active sedimentary basin, and did not significantly vent to surface (McPhie et al., 2011; 2016).
The 'bedded clastic facies' were deformed prior to final lithification, attested to by folds that have diverse shapes and sizes but lack cleavage. These folds have a gentle and reasonably consistent axial plunge to the NE and may be related to slumping on a NE-striking palaeoslope, dipping either to the NW or SE. The depocenter in which they were accumulated is interpreted to have been bounded by a combination of NE- and NW-striking faults (McPhie et al., 2016). Ehrig et al. (2021), after Cherry et al. (2018) show that the age of for the 'bedded clastic facies' tuffaceous mudstone has been constrained to 1590.97 ±0.58 Ma, which implies the probable detritus source, the Roxby Downs Granite, was partially exhumed before ~1591 Ma (Cherry et al., 2018). This is taken to indicate the presence of a basin by ~3 m.y. after emplacement of the granite and felsic volcanic rocks.
McPhie et al. (2016) consider it likely that the 'bedded clastic facies' were deposited in a fault bounded sedimentary basin while the Olympic Dam hydrothermal system was active. As such, that system was probably responsible for sub-sea floor brecciation and alteration within the underlying Roxby Downs Granite and Gawler Range Volcanics basement during deposition of the 'bedded clastic facies' and subsequently progressively stoped upwards to also intersect and breccite the latter. In this sense, the 'bedded clastic facies' are pre- to syn-brecciation.
The 'bedded clastic facies' comprise the following main associations, not necessarily in stratigraphic order:
• KASH - laminated, intercalated green sandstone and mudstone, which is up to 100 m thick and can be traced over a 1.3 km strike length and to a depth of ~750 below the Neoproterozoic unconformity. It comprises interlaminated green sandstone and mudstone, with ~1 m, and occasionally up to 10 m thick green sandstone intercalations, which contain green mudstone intra-clasts. The dominant green colouration of the mudstone is caused by phyllosilicates, mainly chlorite, although fine-grained sericite is also present (McPhie et al., 2016). The sandstone and mudstone generally have sharp contacts, although some sandstone beds grade into mudstone. Laminae and beds are laterally continuous and of even thickness. It is dominantly made up of chlorite-altered mafic volcanic and lesser sericite-altered feldspar-phyric felsic volcanic clasts, as well as chlorite-altered ferromagnesian and sericite-altered feldspar grains that include relict orthoclase and albite, with quartz, carbonate, fine-grained white mica and accessory zircon and titanite. This detritus was derived from mafic volcanic and granitoid rocks. The altered mafic igneous clasts have porphyritic, doleritic and vesicular textures, and a composition suggestive of chlorite-pseudomorphed olivine crystals that contain chromite inclusions and free chromite crystals (McPhie et al. , 2016). As such it is rich in detrital chromite, and is characterised by ~9 wt.% Al, 2.5% Mg, 3.5 to 4.5% K and 0.45 to 0.55% Ti, with elevated 400 to 500 ppm Cr, 100 to 200 ppm V, 100 to 300 ppm Ni and 200 to 400 ppm Zn (Ehrig et al, 2012). Most green sandstones and mudstones sampled also contain more abundant P2O5 (>0.2 wt.%) than the red sediments of the KHEMQ (McPhie et al. , 2016). The sandstone contains interstitial sulphide minerals, including pyrite, chalcopyrite, bornite and chalcocite, together with siderite, barite, fluorite and apatite, whilst minor sulphides are also present within cross-cutting barite ±carbonate veins. Variably brecciation, alteration and subordinate mineralisation is observed throughout the unit, and in addition to the chlorite and sericite, it has experienced hematite alteration where it assumes a reddish-brown colouration. Despite the alteration and presence of sulphides, the mean copper concentration of the unit is only 0.21 wt.% and is therefore largely barren (McPhie et al. , 2016).
However, the distribution of mineralisation is strongly influenced by KASH, whereby increasing copper grades are observed proximal to its faulted contact. A contiguous, albeit narrow zone of high-grade mineralisation occurs along the southernmost faulted contact of KASH, forming a local small extension to the main orebody. This segment of the unit has been significant structurally disturbed, characterised by intense hematite + sericite alteration, with textural destruction of sedimentary features. It is uncertain as to whether this zone is due to post-mineral ‘reworked’ brecciation and remobilisation of copper to oxidized, chalcocite-dominant ore, although it appears likely (Clark and Ehrig, 2019). The >1% Cu mineralisation shown within KASH on Fig. 2 is likely influenced by the proximity of high grade mineralisation near its contact.
• KHEMQ - well-bedded, intercalated, hematite-quartz-rich sandstone and red mudstone, with localised conglomerate-breccias. This unit comprises up to 50% of the 'bedded clastic facies', with preserved thicknesses of up to 90 m. It is composed of hematite, granite and felsic volcanic clasts, and granitoid- and volcanic-derived quartz and feldspar fragments, but lacks significant mafic igneous components. It is mainly composed of red mudstone, which is almost entirely composed of red and grey hematite, with minor magnetite and scattered fine quartz, zircon, white mica and Fe-chlorite. The red mudstone intervals are planar laminated to very thinly bedded. Some mudstone bands are delicately laminated and hematite-rich, with successive laminae of red, grey and black hematite that resemble banded iron formation. The KHEMQ facies association contains anomalous Mo and As of ~150 and >200 ppm respectively, which has been interpreted to represent a hydrothermal overprint. The red mudstone is intercalated with coarse- to fine-grained, thinly, 3 to 10 cm, bedded and laminated white, yellow and pink sandstone. Some thicker, up to 30 cm, sandstone beds and 50 cm thick polymictic, granule to pebble conglomerate and breccia beds are also present. The polymictic conglomerate and breccia beds generally display normal grading upwards into parallel laminated sandstone. Granitoid clasts and granitoid-derived quartz and feldspar are the dominant components of the sandstone, which also contains hematite, zircon, tourmaline, titanite and rutile. The thicker sandstone beds are typically internally massive, parallel laminated or graded, and tabular, although some are lenticular. These sandstones exhibit flame, and ball-and-pillow structures, cross-beds and cross-laminae, defined by dark bands of grey hematite. Quartz, fine-grained white mica and hematite are common sand grains, with interstitial sulphides that include pyrite, chalcopyrite, bornite and chalcocite, with local florencite and Pb telluride (McPhie et al. , 2016).
• VASH - thinly laminated, hematite-rich tuffaceous mudstone, which is a delicately interlaminated green and red, very fine grained (<1 mm), strongly hematite-rich, tuffaceous mudstone facies with minor laminae to thick beds of fine-grained, pale grey, pale brown and purple sandstone. The red mudstone layers are composed of fine-grained hematite and white mica, and the fine-grained sandstone beds contain quartz, feldspar and angular to subangular felsic volcanic clasts. Some of the sandstones are normally graded. Laminae and thicker beds are laterally continuous and of even thickness. Some laminae and thin beds of green mudstone contain sparse chromite grains, and others include magnetite rimmed by hematite. The felsic tuffaceous mudstone contains well preserved bubble-wall shards, fine angular volcanic quartz and zircon. This latter mudstone is pale green, and occurs in internally massive laminae and thin to thick beds (McPhie et al. , 2016).
• KFMU - polymictic conglomerate with felsic and mafic clasts of Gawler Range Volcanics. This unit is only a minor, but locally up to 70 m thick, component of the 'bedded clastic facies', and is in faulted contact with the VASH unit. The predominantly conglomerate facies is interbedded with structurally underlying green sandstone and mudstone and an overlying interbedded hematite-rich sandstone and red mudstone facies. It is distinguished by its much coarser grain size, with most clasts being 1 to 6 cm in diameter. The conglomerate is characterised by pink to red hematite- and green chlorite-altered volcanic cobbles and pebbles. Most clasts are feldspar-phyric and felsic in composition, and very similar to feldspar-phyric rhyolites of the Gawler Range Volcanics. Less than 1% of the felsic volcanic clasts are quartz-phyric. Some clasts are mafic in composition, reflected by pseudomorphs of skeletal olivine phenocrysts that contain chromite. Granite and minor hematite clasts are also evident. Most clasts are well rounded, although some are angular. It is typically clast-supported, but locally matrix-supported. The matrix is a sandstone composed of polymictic volcanic fragments with lesser feldspar, chromite and rare quartz and pseudomorphs of olivine. The chlorite- or hematite-altered matrix has a patchy green or purple colour respectively (McPhie et al. , 2016).
• VBx - polymictic volcaniclastic breccia, which is a hematite-rich breccia composed of porphyritic felsic volcanic clasts (Ehrig et al., 2012), although Ehrig and Clark (2020), Ehrig et al. (2022) and Corriveau et al. (2022) state that it contains both felsic and mafic-to ultramafic clasts from the Gawler Range Volcanics, all set in a 'sea of hematite' matrix.
The first three of these facies, namely KASH, KHEMQ and VASH, are interpreted to have been deposited in a predominantly low-energy, below wave base, subaqueous environment (McPhie et al., 2011), whereas KFMU was laid down during a higher energy depositional event. All of the 'bedded clastic facies' have textures that are indicative of local soft sediment slumping prior to final lithification. Zircon grains from the 'bedded clastic facies' have age populations around 1600 Ma (McPhie et al., 2016). It is considered likely (McPhie et al., 2016) that all of these facies were deposited in a fault bounded sedimentary basin that was active when the breccia complex formed and the hydrothermal system was active.
A further clastic facies has been encountered, but as of 2017, had only been intersected in two drill holes. It is an unaltered, well-sorted, quartz-rich sandstone with faulted contacts and contains zircon grains dated at ~1590, ~1740, ~1900 and ~2500 Ma. This zircon age spectrum is identical to that of the Pandurra Formation, implying that the breccia complex was reactivated, or deformed, after 1424 Ma, the diagenetic age of the Pandurra Formation (Ehrig et al., 2017). However, it seems likely that sections of the Pandurra Formation may be as old as 1575 Ma. This intersection is also predominantly brecciated, with a hematitic sand matrix as described below, but probably post-dates the main Olympic Dam Breccia Complex event.

Alkaline ultramafic to mafic lavas, dykes and possible sills comprise ~30 vol.% of the Olympic Dam Breccia Complex, including those interpreted as the protoliths of heavily hematite altered rocks (Ehrig, 2022, from unpublished mass balance calculations). These include:
Ultramafic volcanic units, occurring as remnants of what are now intensely altered chromite-bearing, olivine-phyric picrite. They are found immediately below the unconformity with the Neoproterozoic cover sequence in the southeastern quadrant of the deposit, adjacent to the 'bedded clastic facies' occurrences. These remnants vary from 15 to 45 m in thickness in eight drill holes, and may represent either a single, or multiple units. The picrite is juxtaposed with various hematite-rich breccias, separated by intensely altered and sheared contacts. Whilst it is strongly carbonate-hematite altered, it has a well-preserved porphyritic texture comprising evenly distributed, skeletal, ~1 mm diameter, relict olivine phenocrysts, and up to ~1 cm wide carbonate-hematite filled amygdales. Carbonate-hematite veins that are 1 to 5 mm in width are also present. Additional diagnostic properties of picrites include elevated concentrations of >0.1% Cr, 100 to 200 ppm V and 300 to 500 ppm Ni. On the basis of its location and texture, this picrite is interpreted to have been a lava (Ehrig et al., 2012).
Pre- to Syn-brecciation Dykes. A suite of mafic and felsic igneous intrusions, mainly dykes, are interpreted to have been emplaced just prior to or during brecciation and mineralisation. The mafic dykes are generally steeply dipping, intruding the Roxby Downs Granite, 'bedded clastic facies', and Olympic Dam Breccia Complex breccias. Some are mineralised, and are both syn-breccia and syn-mineralisation. Multiple generations of dykes commonly intrude along the same structures. Four types are distinguished, based on composition and alteration (after Ehrig et al., 2012), namely:
• Aphanitic, sericite-hematite-altered, probably mafic to ultramafic, dykes, which have been altered to microcrystalline sericite and chlorite and commonly hematite. These dykes, which are best exposed in the deeper parts of the northwestern ore zone, vary from <0.5 up to >40 m in thickness and have inferred strike lengths of as much as 1 km. The contacts with granite-rich breccias are sharp at depth and become more gradational with decreasing depth. They contain micron sized sericite and crystals of zircon, rutile, REE-bearing minerals xenotime, bastnäsite and florencite, and have a characteristic 'waxy' sericite feel/texture. As such, they are anomalous in Ti (>0.3 wt.%), Mg (13 wt.%), Zr (>400 ppm), Al (>9 wt.%) as well as containing other HFSE (e.g., Hf, Th, Ta, Nb), U and K (Ehrig et al, 2022). Intense hematite alteration is common, in which case they are mapped as HMBx. High grade sulphide concentrations are often associated with the margins of, or within these dykes (see Fig. 2 and cross section to the NW of the Masher's Fault on Fig. 3). The chemical composition is interpreted to suggest a mafic to ultramafic protolith, possibly lamprophyre (Ehrig, 2022). They appear to be closely associated with the introduction of mineralisation.
• Sericite-altered, olivine-phyric mafic dykes, which although generally strongly altered, retain a relict porphyritic texture as pseudomorphs after olivine phenocrysts in a very fine groundmass. Olivine appears to have been the only phenocryst in the protolith, and its strong abundance suggests these dykes may have been ultramafic in composition. They are a few to a few tens of centimetres thick and have sharp contacts with the granite and more irregular, lobate or brecciated contacts with the hematite-rich breccia. The contacts are marked by ~3 mm wide glassy zones within the dyke, and ~5 mm wide oxidised rims in the granite.
• Sericite ±chlorite ±hematite-altered, plagioclase-pyroxene-phyric basaltic dykes, which range from relatively fresh to strongly chlorite ±sericite ±hematite altered. Phenocrysts are generally 1 to 2 mm across and difficult to differentiate from the groundmass when strongly altered. These dykes are generally steeply dipping, and at 1 to 5 m thick are relatively narrow, although locally widening to tens of metres. They may occur as cross-cutting, green chlorite-rich dykes that also contain disseminated pyrite and chalcopyrite, and veins of chalcopyrite, pyrite with fluorite and barite. They have a similar composition and texture to the Gairdner Dykes.
• One intensely altered, quartz-phyric felsic dyke has been intersected at a depth of ~1900 to 2300 m in the deepest vertical drill hole (RD2773) in the southeastern part of the Olympic Dam Breccia Complex. It is composed of melt inclusion-bearing quartz phenocrysts in a groundmass of quartz and sericite with minor zircon and apatite, interpreted to imply an original rhyolitic composition. It intrudes weakly brecciated, chlorite-altered Roxby Downs Granite, has brecciated margins and is steeply dipping (Ehrig et al., 2012). Preliminary laser ablation inductively coupled plasma mass spectrometry (LAICP-MS) zircon dating returned an age of ~1590 Ma. On the basis of the intensity of alteration and the contact relationship with the Roxby Downs Granite, it is interpreted to probably be slightly older than the granite (Ehrig et al., 2017).
Johnson and McCulloch, (1995) estimated the ultramafic to mafic rocks accounted for >13 vol.% (~30 vol.%; Ehrig, 2022) of the breccia complex, but contained ~50% of the copper in that complex. Huang (2016) demonstrated that these ultramafic to mafic rocks contain extreme concentrations of some components (e.g. up to 26 wt.% CO2, 50 wt.% of Fe2O3 and 6000 ppm Cr) and 'extraordinary' near linear positive correlations between Cr and high field strength elements such as Ti, Nb and Zr, which is taken to indicate significant whole-rock mass and/or volume loss due to hydrothermal alteration and hence a close association with mineralisation. See the Mineralisation section below for more detail on the role of these mafic to ultramafic dykes.

ODBC - Olympic Dam Breccia Complex
The ~50 km² and 500 to >1000 m thick Olympic Dam Breccia Complex was principally developed within the Roxby Downs Granite, and to a lesser extent within the Gawler Range Volcanics and bedded clastic facies, with contributions from hydrothermal processes. The degree of alteration and intensity of brecciation increases from the edges of the complex toward the centre of the deposit, with the breccia composition and textural characteristics being gradational. As such, most rocks within the Olympic Dam Breccia Complex form a continuum, ranging from weakly sericite-altered Roxby Downs Granite, to sericite-altered granite-rich breccias, through to hematite-dominant breccias in the deposit centre. The subdivision of the breccias of the complex, based on primary rocks is complicated by the intense hydrothermal alteration, although on a deposit scale, compositional changes across the breccia continuum are systematic and symmetric. The breccias are classified according to the total concentration of iron oxide minerals with clast composition qualifiers. Breccias range from <5 to >90% iron oxides (Ehrig, McPhie and Kamenetsky, 2012) with corresponding whole-rock iron concentrations of up to ~68%.
The texture, contact relationships, distribution and non-stratified nature of the breccias of the complex are interpreted to be consistent with subsurface fragmentation by a combination of tectonic and hydrothermal processes (Oreskes and Einaudi, 1990; Reeve et al., 1990). Further, angular or ragged, amoeboid, and digitate clast shapes suggest both brittle failure and dissolution of preexisting wall rock were important in the generation of these breccias (McPhie et al., 2011). Ehrig and Clark (2020) describe the evolution of the breccia as commencing on the periphery of the complex as granite with fractures paralleling the NE-SW and NNW-SSE fault pattern (see Fig. 1 above and Structure section below) on a metre to tens of cm spacing, forming fracture bounded rhomboidal blocks of granite. Stepping in towards the deposit, the granite becomes more fragmented with closer spaced fracturing. These fractures are the focus of hematite-sericite alteration, replacing the margins of the rhomboidal granitic blocks and rounding their angular corners at fracture intersections. As this process progresses, the granite blocks become more intensely fractured, and are reduced in size by hematite-sericite alteration replacing the granitic blocks/clasts inwards towards their core. This produces a replacive breccia with sub-angular clasts in a hematite-sericite matrix. As the brecciation and alteration progresses, and the amount of matrix increases, clasts from higher in the sequence, such as the Gawler Range Volcanics and 'bedded clastic facies' are found in the breccia, suggesting the highly corrosive fluorine-rich hydrothermal fluids have dissolved and reduced the volume within the breccia complex, allowing 'solution-collapse' and formation of heterolithic breccia. Towards the centre of the deposit, higher grade Cu mineralisation and fluorine content prevail, where the hydrothermal processes have dominated the initial physical fault controlled brecciation, the hematised granite clasts and remnant quartz grains are sub-rounded to rounded, with the latter being cloudy, while many of the clasts are composed of other hematised lithologies and multiple generations of breccia and hematite.
The principal breccias developed within the Roxby Downs Granite have been divided on the basis of Fe content by Ehrig (2019, 2021, 2022), from the periphery (and at depth) to the upper core of the complex, into:
i). GRN - altered, weakly brecciated granite, to granite-rich breccia, which is biotite-free, and contains <5 wt.% Fe;
ii). GRBx - granite, to hematite-rich breccia, with 5 to 25 wt.% Fe;
iii). HMBx - hematite-rich breccia, with >25 wt.% Fe;
iv). HEMQ - hematite-quartz-barite breccia, with >60 wt.% Fe;
The geological description of the transition these divisions represent was best described by Reeve et al. (1990), a summary of which is as follows:

Brecciated Granite to Granite-Rich Breccias, which span a wide range of textures that reflect variable intensities and styles of brecciation and fragment transport, ranging from sparsely fractured or veined granite on the margins of the complex, to coarse clast supported breccia with rotation of blocks/clasts inward. Crackle and jig-saw textures are commonly preserved in matrix poor granite breccias, particularly in less altered zones. These textures become progressively less recognisable as the proportion of matrix increases. Apart from superimposed alteration, the clast lithology is indistinguishable from Roxby Downs Granite external to the complex. This transition is interpreted to comprise a replacive breccia, where the fracture controlled veining/alteration expands, consuming the granite wall-rocks which remain as progressively smaller preserved kernels, and when isolated in the alteration matrix may rotate.
Some of the granite-rich breccias are composed of isolated large clasts of granite that are up to tens of cm across, supported by a finer grained, variably altered, granite-derived matrix. In these breccias, the matrix may show some sorting, with locally developed discontinuous and diffuse layering, and some local, irregular, matrix-rich micro-breccia patches resembling arkoses or diamictite-like fabrics. The bulk of the clasts and matrix fragments are angular to subangular, and subequant to elongate in shape, with only rare instances of substantial rounding in either. However, where there is an intense sericitisation or silicification alteration overprint, clast boundaries are generally highly irregular or diffuse.
Breccias with granite fragments within a hematite rich matrix are part of the continuum to hematite-rich breccias and locally contain abundant sulphides. These breccias range from matrix poor granite breccias containing hematite rich crackle veins, through more matrix rich transitional varieties, to breccias dominated by a hematite-quartz ±sericite/muscovite matrix containing minor barite and fluorite. Some contain cataclastic fabrics, whilst others have unusual fabrics suggestive of both ablation and attrition.

Altered granite breccia

Hematite-Rich Breccias are extensively developed, and the most important host to mineralisation. These have been subdivided into three broad, transitional breccia types, as described individually below. However, all three have common textural features, as follows:
i). fabrics range from matrix to clast rich, although the former are generally more abundant;
ii). the clast size is very variable, and although isolated blocks up to metres, or rarely tens of metres in diameter occur, most are generally <20 cm;
iii). although local zones of limited sorting have been observed, most breccias are poorly sorted;
iv). clasts may be locally predominantly irregularly shaped, but are usually angular and approximately equi-dimensional;
v). substantial rounding of clasts is rare, even where the presence of heterolithic breccias suggest remobilisation and mixing;
vi). repeated brecciation is indicated in many of the breccia clasts and bodies.
The spectrum of hematite-rich breccias includes:
• Heterolithic Hematitic Breccia - a broad grouping that includes the greatest proportion of hematite-rich breccias at Olympic Dam, and hosts the bulk of the Mineral Resource. It is predominantly composed of a range of hematite clasts, but may also contain a significant proportion of altered granite fragments, with the amount of each clast type being highly variable. Those clasts that are mostly hematite-rich, range from fine-grained and dark red-brown, to fine- or medium-grained crystalline black varieties. Less commonly, these clasts include hematite that has a coarse feathery-texture and is crystalline, brick-red and black powdery, or dense steely grey. The strongly heterolithic nature of these breccias suggest they have been subjected to a degree of mixing, and possibly significant transport. Never-the-less, these heterolithic hematitic breccias may contain poorly defined 'pockets' that are metres to tens of metres across where similar clast types dominate, or in which the granite clasts have the same composition and alteration type. Granite clasts of conspicuously different alteration style are rarely found in very close proximity.
Other clast lithologies, though minor, contribute to the heterolithic nature of these breccias, and include highly altered, fine grained, mafic and felsic igneous intrusive material, finely laminated sedimentary rocks that are predominantly composed of hematite and quartz, and massive to diffusely layered medium to coarse grained arkose-like rocks containing altered granite-derived feldspar and quartz fragments. Some of these breccias also contain minor clasts of massive chalcopyrite, barite, fluorite or siderite. Localised breccia zones near the northeastern margin of the deposit include a major proportion of siderite clasts.
The matrix of the heterolithic hematitic breccias is normally dark red-brown to black, fine- to medium-grained hematite, with a variable amount of quartz and other material largely derived from granite. Commonly there is an irregular 'streaky' mixture of dark red-brown and black hematite that may define a crude layering. Some of these layers may also have undergone limited sorting, particularly among the non-hematite grains. Copper sulphides, pyrite, fluorite and barite may also be disseminated in the matrix.

Heterolithic hematite breccia

• Hematite Breccia, in which both the clasts and matrix are largely composed of hematite. Although this breccia type is the least abundant of the 'hematite-rich breccias', it may be the most strongly mineralised, preferentially containing the highest grade ore, particularly in the northwestern parts of the deposit. However, where developed well below the bornite-chalcopyrite interface, lower in the mineralised system, it commonly contains abundant pyrite and only minor copper sulphides.
Hematite breccias display a wide variation in colour, texture, and composition of the minor non-hematite component. The breccia is generally dark grey to black, reflecting the predominance of hematite grains with well developed crystal forms. They are composed of fine- to medium-grained, obvious crystalline hematite with interlocking mosaic, tabular or bladed textures, and have a distinctly contrasting appearance to the other two hematite-rich breccia types in which pulverulent, earthy or dusty red-brown to purplish hematite is more common. These black to grey hematite breccias range from dense and hard, to very porous or vuggy types, with extremely porous varieties containing up to ~30% voids. Some hematite breccias contain abundant small vugs which appear to be the result of leaching of sulphides. Some are only weakly brecciated, or appear to have been brecciated in situ, although in some cases, there is only very subtle textural differences between the clasts and matrix and they are difficult to discern.
Non-hematite components are variable, and may include quartz, chlorite, fluorite, barite, altered granite clasts and granite-derived mineral fragments, siderite, minor relict magnetite and sulphides ±pitchblende in mineralised areas. Reeve et al. (1990) note that minor amounts of hematite breccia may contain distinctive components and/or exhibit unusual textures, including hematite-rich or hematite-sulphide pisolites, strongly laminated hematite rich rock fragments, wispy discontinuous layering, crustiform hematite, and coarse feathery-textured hematite. These may occur as individual isolated clasts within hematite breccias, or as small 'pockets' of particular clast types.

Hematite breccia with sulphides

• HEMQ - Hematite-Quartz Breccia is the most intensely altered rock type within the Olympic Dam Breccia Complex. It is composed of clasts of hematite, quartz and hematite-quartz set within a matrix that is predominantly hematite with grains of quartz. It is texturally destructive and resulted in Al depletion corresponding to the removal of all aluminosilicate minerals. It is largely restricted to the shallow central core of the breccia complex, although minor amounts of similar breccias are found elsewhere within the deposit. This breccia, which is predominantly in the hanging wall of the ore, is characterised by abundant quartz fragments, particularly in the matrix, where they may constitute as much as 40% of the rock. Reeve et al. (1990) note that the crystal morphology and characteristic undulose extinction of this quartz indicates it was derived from the local granitic lithologies, although fragments of vein quartz and/or a range of other strongly siliceous or silicified rocks are locally abundant. However, Ehrig et al. (2021) suggest the bulk of the HEMQ is after lithologies of the Bedded Clastic Facies.
The HEMQ typically contains ≥110 ppm As, <2.5 wt.% Al and < 0.35 wt.% Cu (Ehrig and Clark, 2020).
The Hematite-quartz breccia is also characterised by dominant fine grained dusty red-brown to dark purplish hematite in both clasts and matrix. Steel-grey to black hematite clasts are also present, although much less abundant and generally smaller than the coexisting red-brown or purplish types. It is also noted for the abundance of barite, generally 2 to 5%, occurring as disseminations, veins and vein fragments, and for the scarcity of fluorite compared to other hematite-rich breccias.

Post-brecciation Rock Units
KGRN - Pandurra Formation, which regionally which was deposited in the ∼42 500 km², ~420 x 170 km Cariewerloo red bed basin that unconformably overlies the Gawler Range Volcanics, as described above in the Regional Setting section and the Gawler Craton and Olympic IOCG Province record. However, as described above, two drill holes (as of 2022), have each intersected ~80 m an unaltered, well-sorted quartz-rich sandstone with faulted contacts which contains a spectrum of detrital zircons identical to those of the Pandurra Formation. It has therefore been interpreted to represent a slab of Pandurra Formation overlying the 'bedded clastic facies' which has been structurally interleaved with the latter and down-faulted into the Olympic Dam Breccia Complex (Ehrig et al., 2017). This sandstone is almost entirely brecciated, with only the uppermost 2.3 m and another lower ~2 m section that appear to be un-brecciated, although both could conceivably be large clasts. The breccia clasts vary from <1 to tens of cm across and range from angular jigsaw fit to separated sub-rounded blocks. They are composed of well sorted, quartz-rich, massive to diffusely stratified sandstone, with some being cross-bedded. The sandstone of the clasts is composed of sub-rounded quartz and mm-sized muscovite grains, with tourmaline and fine-grained red hematitic fragments. The quartz grains have undulose extinction and abundant fluid inclusions, whilst some are poly-grain aggregates with annealed, triple-junction boundaries. The breccia clasts are set in a poorly sorted hematitic matrix that is markedly different in character to the clast lithology, but comparable to the hematite breccias. This brecciated sandstone has a sharp upper and gradational lower contact with polymictic hematitic breccia of the Olympic Dam Breccia Complex. No other drill holes on the same section intersect this unit, with the nearest passing 50 and 250 m to the SW and NE respectively (McPhie et al., 2016).
Gairdner Dykes, which have been dated at between 827 ±6 and 824 ±4 Ma, and are weakly altered, varying from coarse, equigranular dolerite to finely porphyritic basalt. They are composed of equigranular pyroxene and plagioclase crystals that are intergrown in the dolerite, or occur as phenocrysts in the porphyritic basalt. Magnetite is responsible for their characteristic magnetic susceptibility of ~>100 x 10^-3 SI units. Multiple generations of dykes are evident in some cases, with early dolerite being intruded by later porphyritic basalt. Contacts between these dykes and the Roxby Downs Granite or Olympic Dam Breccia Complex are sharp and marked by perlite or palagonite-altered glassy margins typically a few millimetres thick. In some instances, the dolerite may be weakly chlorite altered with incipient hematite alteration at shallower levels, whilst narrow carbonate veins are common. Although weakly altered, this dolerite is not mineralised. It is characterised by Cr, V, Ni and Ti contents of 50 to 100 ppm, 300 ppm, 50 to 100 ppm and 1.5 to 2 wt.%, respectively.
Neoproterozoic to Cambrian sedimentary rocks, that form the cover sequence that comprises ~350 m of flat-lying, shallow marine shale, quartzite, and dolostone, unconformably overlying the sub-horizontal eroded upper surface above the Roxby Downs Granite, Olympic Dam Breccia Complex and Mesoproterozoic sedimentary sequences.

STRUCTURE

A complex array of brittle faults are seen within and surrounding the deposit. There are interpreted to include pre-, syn- and post-mineralisation structures (Clark, Passmore and Poznik, 2017). Ehrig et al. (2017), however, note that annealing by intense alteration overprinting of pre- and syn-mineralisation structures hampers the easy recognition of these faults. Never-the-less, the same authors suggest the NW strike of ore zones in the north-western part of the deposit, sharp geochemical fronts and evidence of multiple brittle deformation events within the breccias are strong indicators of pre- and syn-mineralisation structures. Post-mineralisation faults are evident in the central section of the deposit where they disrupted primary mineralisation trends. In addition to numerous local-scale structures, five main fault sets are recognised in and around the deposit, all of which are steeply dipping. However, rotation related to late stage faulting may have changed the orientation of earlier structures. It should also be noted that some of the earlier trends are not obvious on Fig. 1 as they are not major structures. Ehrig et al. (2012) described each of the fault sets as follows:
• North-south faults, that are generally discontinuous and offset by NW, NE and WNW sets, although there is some suggestion of late reactivation. The longest mapped faults of this set extends for over 1 km. Enrichment of hematite and copper is associated with this trend, whilst 'bedded clastic facies' blocks and pre- to syn-mineralisation dykes are located at the intersection with WNW faults.
• NW-SE to NNW-SSE faults reflect a trend that is regionally continuous (e.g., the Elizabeth Creek fault zone of Hayward and Skirrow, 2010 which is ~40 km to the SW). These faults are offset by NE structures and by Masher's Fault. Within the deposit, these structures are associated with hematite alteration, copper mineralisation, pre- to syn-mineralisation mafic dykes and late post-mineralisation dolerite intrusions. There is some evidence of late reactivation.
• WNW-ESE to east-west faults, that are regionally discontinuous and appear to splay from NW faults in the vicinity of NE structures, although they may also be slightly earlier than the NW faults. Hematite alteration, copper, and locally molybdenum mineralisation are associated with these structures where they occur in the immediate hanging wall of NE faults. There is evidence of minor reactivation, which is intensified closer to Masher’s fault with sinistral offsets. In the deposit area, this trend is represented by the Woodall Fault and Woodall Fault Junior, ~400 m to its north, which both dip steeply to the NNE. These faults together, have been estimated to represent a 2.5 to 3 km of north-side down normal offset. They have been responsible for the relative uplift of the Roxby Downs Granite and downthrow and tilting to near vertical of the 'bedded clastic facies', and probably persisted until after the deposition of the Pandurra Formation on its northern side. This displacement brings the unaltered Roxby Downs Granite in much closer proximity to the centre of the deposit than elsewhere. This downthrow has also been responsible for the deepest occurrence of mineralisation, with drill holes terminating in grades of >2% Cu and 1.5 g/t Au at depths of >1750 m depth (Ehrig and Clark, 2020). Abundant 1 to 5 mm thick quartz-tourmaline, quartz-carbonate and quartz dominant veins, which are rare elsewhere within the deposit, do occur between the Woodall and Woodall Jr faults. In addition, both ~1590 Ma mafic to ultramafic and ~825 Ma Gairdner dykes intrude along both of these faults suggesting long lived activity from at least 1590 to 1424 Ma. Ehrig (2022) queries whether the Woodall and Woodall Junior faults are an annealed 'mega-shear' as although displacement is indicated, no well defined fault plane has been intersected in drilling.
• NE-SW to ENE-WSW faults, which are also regionally continuous (e.g., the Todd Dam fault zone; Hayward and Skirrow, 2010) with the most recent displacement being sinistral. They offset NW, WNW and north-south fault sets. Some are markedly vertical and penetrative, and control the emplacement of mafic dykes. Others are now listric, e.g., the Oreskes Fault, which represents an original major domain boundary that was a normal fault during development of the deposit. Where NE-SW and WNW faults intersect, they are associated with both hematite alteration and copper mineralisation. Regionally they are mapped as being reverse structures, although at the deposit scale, there is evidence to suggests normal displacement. Where they are steeply dipping, they have been strongly reactivated in association with the sinistral movement on the major Masher's fault, described below, to which they are parallel to sub-parallel.
The prevailing structural architecture in the southeastern lobe of the deposit is predominantly composed of a series of east- and west-dipping normal faults with a NE-SW orientation. These faults splay from, or displace, the Woodall and Woodall Junior faults with normal offsets of tens to hundreds of metres. Together with the Woodall Junior and Lalor faults to the south and north respectively, these faults fundamentally controlled the initial structural dismemberment of the deposit and incorporation of the allochthonous fault dislocated, ~1.5 x 0.6 x 2 km block of 'bedded clastic facies' into the Olympic Dam Breccia Complex (Fig. 1; Clark and Ehrig, 2019).
• Listric extensional faulting is evident within the deposit area, occurring as moderate to shallow to flat lying normal faults (Fig.s 2 and 3) that mark some lithological boundaries (e.g., the base of the HEMQ breccia on Fig. 3). These structures appear to be mostly post-mineral, with many being truncated by late steep NE-SW and north-south structures, although others (Fig. 2) appear to be the down-dip sections of some of the NE-SW faults that are steeper at surface. However, some of these listric faults appear to parallel shallow dipping breccia zones and high grade mineralised ore shoots, and may represent reactivated earlier extensional structures.
• Masher's fault, the local mine name for the regional Jubilee fault zone (Hayward and Skirrow, 2010) which is a major zone of crustal weakness. Regionally it has an ENE trend, and sinistral displacement, but in the deposit area parallels the NE-SW trending structures and is a 150 m wide zone of faulting. During the history of the Olympic Dam deposit it has been reactivated on several occasions. Typically, whenever there has been a major deformation event within the Gawler Craton, hydrothermal modification and structural dislocation is recorded in the deposit, centred on the Masher's Fault zone, and from there, distributed through its subsidiary structures. The main body of the deposit is centred on the Masher's Fault intersection with the postulated pre-mineral NNW-SSE (to NW-SE) fault that is inferred to control the main elongated mineralised zone to the NW. It does not appear to have influenced the original development of hematite breccias or copper mineralisation, but does markedly truncate and offset both lithofacies, alteration and mineralisation. In the deposit area, south side down normal displacement has preserved substantial fault blocks of 'bedded clastic facies' in the southeastern lobe of the deposit, juxtaposed against 'deeper' brecciated and mineralised Roxby Downs Granite to the NW of the fault. It was active as late as the Cambrian, displacing the Early Cambian Andamooka Limestone in the overlying cover sequence and is reflected by a ~100 km long regional aeromagnetic feature.

ALTERATION

Skirrow (2022) notes that regionally, Na-Ca-Fe metasomatism is represented in the Olympic IOCG Province by assemblages of hydrothermal albite-calcsilicate comprising actinolite and/or diopside ±magnetite, commonly accompanied by minor titanite and scapolite. This alteration typically forms kilometre-scale regional alteration zones in both the Mount Woods and Moonta-Wallaroo districts, well to the north and south of Olympic Dam respectively. However, sodic alteration is much less evident in the Olympic Dam district where rarely observed hydrothermal albite is suspected to represent relicts of sodic metasomatism surviving after widespread K feldspar overprinting (Bastrakov et al., 2007). This Fe²⁺-K metasomatism is represented by magnetite-K feldspar alteration. Relative timing relationships indicate that regionally, albite-actinolite ±magnetite alteration is overprinted by a biotite-magnetite assemblage, which, in turn, is replaced and cut by the hematite-sericite-chlorite-carbonate alteration typical of Olympic Dam (Skirrow, 2022).
The principal hydrothermal alteration minerals at Olympic Dam include hematite, magnetite, muscovite (1M and 2M1 types) and quartz, with lesser chlorite (chamosite/chlinoclore), siderite, fluorite, barite, paragonite, albite and K feldspar (Oreskes and Einaudi, 1990; Reeve et al., 1990; Ehrig, McPhie and Karnenetsky, 2012; Mauger et al., 2016; Kontonikas-Charos et al., 2017; Ehrig et al., 2017).
Iron oxide alteration represents: i). a progressive spatial increase in Fe from edge and depth to centre and top; and ii). a temporal shift from the more reduced Fe²⁺ → the more oxidised Fe³⁺, represented by magnetite + apatite + chlorite → hematite + sericite. Another zonation from edge to upper-core is represented by siderite → fluorite → barite.
There is a complex distribution of alteration minerals throughout the hydrothermal system, although deposit-scale alteration assemblage trends are evident. Beginning from the periphery, alteration is first seen in the Roxby Downs Granite, distal to the Olympic Dam Breccia Complex. The outer margin of the breccia complex is marked by the boundary between the external 'unaltered' and the weakly altered Roxby Downs Granite. This boundary has been defined to be at the 'biotite-out' contour, internal to which no magmatic biotite remains. However, this alteration change, which is 1 to 5 km from the deposit centre, is transitional, with external incipient sericite/muscovite alteration of plagioclase, while igneous biotite is progressively converted to chlorite, until no biotite remains at the 'biotite-out' contour.
Alteration then progressively increases in intensity inward within the Olympic Dam Breccia Complex, where iron metasomatism is much higher and hydrothermal K feldspar is replaced by muscovite.

Three major deposit-wide zoned alteration assemblages and one minor style are recognised (Reeve et al., 1990; Ehrig, McPhie and Kamenetsky, 2012; Ehrig et al., 2017; Ehrig and Clark, 2019; Corriveau et al., 2022), namely:
i). Paragenetically early, relatively reduced, Fe oxide alteration, characterised by high temperature (~800 to 400°C) magnetite-fluorapatite-ankerite-calcite/siderite-chlorite-quartz, which occurs at depth and outward in the deposit on the periphery of the mineralised areas with ~5 wt.% Fe. This early alteration, referred to as the 'footwall distal alteration' assemblage, involved the development of a high temperature, iron-rich, Ca-Fe alteration (i.e., an early IOA style mineralisation) that evolved to high ƒCO2 conditions, low calcium contents and crystallisation of siderite and chlorite. This zone has an elevated magnetic susceptibility compared to the rest of the deposit, reflecting the relict magnetite being progressively overprinted by siderite/ankerite and later by hematite inwards and upwards into the mineralised system. It also corresponds to the chalcopyrite-pyrite mineralised zone which typically has <1% Cu, as described below in the Mineralisation section. This transition is reflected by chalcopyrite-pyrite mineralisation cutting early high temperature Ca-Fe magnetite-apatite alteration. This is subsequently brecciated and infilled by calcite, while being cut by chlorite-sericite-chalcopyrite and pyrite, as preserved in the deeper, peripheral part of the Olympic Dam Breccia Complex. The siderite and then hematite (after magnetite) formed as the hydrothermal system evolved with a chalcopyrite-pyrite mineral assemblage overprinting the earlier high temperature Ca-Fe facies. Hyper-saturation of CO2 led to an influx of siderite-stable alteration, accompanied by hematite alteration of earlier high temperature Ca-Fe facies and deeper calcite alteration. Siderite also occurs as hydrothermal infill and cement within syn-mineralisation fault zones.
Drilling on the eastern margin of the complex, external to the mineral resource, to a depth of ~2330 m, encountered this assemblage with minor disseminated chalcopyrite, sericite, chlorite, rare earth element-fluorcarbonates, monazite, uraninite, thorite, galena, sphalerite, anhydrite, schorl, rutile and pyrrhotite, cut by abundant multiphase veinlets and calcite ±fluorite ±barite veins (e.g., Apukhtina et al., 2017). Ehrig et al. (2017; 2010) show siderite occurring as a well defined zone at depth and as a 'patchy' halo surrounding the main deposit at shallower levels, overlapping and passing inward into the zone of greater fluorite concentration. The siderite also closely corresponds to the distribution of K feldspar, although it is uncertain from the literature how much of the latter is an alteration assemblage, as distinct from original magmatic feldspar in the Roxby Downs Granite that has not been altered to chlorite and/or muscovite. The siderite and K feldspar appear to underlie/flank the high grade Cu and U3O8, but overlap a substantial amount of the lower grade underlying and outer peripheries of that mineralisation.
ii). The Fe²⁺ stable early siderite/ankerite and magnetite-bearing assemblage is replaced by a low temperature, more oxidised Fe³⁺ K-Fe and Ca-Fe facies, characterised by the assemblage of hematite-muscovite-quartz-fluorite. This alteration assemblage is referred to as the 'ore zone alteration'. The transition is typically sharp and associated with a marked increase in sulphide content, including bornite-chalcocite mineralisation. This assemblage progresses from the poorly brecciated, weakly hematite altered Roxby Downs Granite overlapping the siderite/magnetite zone along the margins of the Olympic Dam Breccia Complex, through increasing muscovite-hematite, into intense hematite-muscovite, to dominantly hematite alteration. This hematite-rich assemblage is the principal host to the economic sulphide mineralisation, with the Fe²⁺/Fe³⁺ redox boundary broadly coinciding with a sulphide mineral zoning transition from chalcopyrite to bornite-chalcocite assemblages. The latter transition varies from being 'knife-sharp' to tens of metres wide. This boundary also corresponds to an increase in fluorite which accompanies most of the bornite-chalcocite mineralisation. The latter constitutes the bulk of the high grade ore which also straddles the transition outward into the chalcopyrite zone (Ehrig and Clark, 2020).
An inner marginal zone of muscovite, illite ±hematite ±fluorite alteration is enriched in gold and lead, which is interpreted to represent later, post-mineralisation remobilisation, spatially controlled by second- and third-order structures in favourable alteration-protolith sites (Corriveau et al., 2022).
iii). An end member assemblage of hematite-quartz-barite, with no sulphides is formed by the complete conversion of muscovite to hematite. This includes the replacement of all aluminosilicates (including chlorite and muscovite) in the upper core of the deposit, and in fault limited smaller zones along the north-west arm of the deposit. This is an end-member alteration type where the original protolith texture has beeen almost totally obliterated, and virtually completely replaced by hematite. This stage includes Ba-Si-Fe and Si-Fe facies, both of which are deposited at low temperature, and is referred to as ‘hanging wall-type alteration'. This assemblage reflects another redox boundary between S^1- sulphide bearing assemblages to SO₄^2- sulphate stability. Hematite and quartz occur as clasts within a matrix of hematite and quartz grains, with minor fragments of vein quartz and/or a range of other high siliceous or silicified rocks being abundant locally (Reeve et al., 1990). Barite occurs as disseminations, vein fragments and abundant narrow, <1 mm barite veins with barium averaging ~2 to 5 wt.% (Reeve et al., 1990). Sulphide mineralisation, which is rarely >0.05 wt.%, is confined to minor late veins and joints. This alteration assemblage has margins that are commonly silicified and barite veined, and may contain gold-only mineralisation (Ehrig, McPhie and Kamenetsky, 2012). Variations and ghosts in the texture may reflect different, extremely altered protoliths (e.g., Roxby Downs Granite, felsic and mafic volcanic fragments after Gawler Range Volcanics, and 'bedded clastic facies'; Clarke and Ehrig, 2019). This alteration assemblage is the HEMQ breccia (described above in the Geology section), which is barren of Cu and U, but may contain Au on its outer margins. It is largely confined to the upper core of the deposit, although other more restricted zones are distributed along the dominant NW-SE axis of alteration and mineralisation (Reeve et al., 1990; Ehrig, McPhie and Kamenetsky, 2012). Its lateral and basal margins are typically faulted, and are interpreted to be preserved as down-faulted blocks spatially associated with the bedded clastic facies. The HEMQ is up to 700 m thick, and extends for from 400 to 750 m below the Neoproterozoic unconformity.
Contiguous zones of high-grade mineralisation follow the margins of the HEMQ. Although the geometry of the HEMQ varies significantly along its strike length, it appears to critically control the distribution of much of the mineralisation, which, if present, wraps around the contact from where it is steep on its lateral margins, to flat-lying beneath the base of the main 'bowl-shaped' HEMQ mass (Figs. 2 and 3; Clark and Ehrig, 2019). Mineralisation is dominantly chalcocite-bornite proximal to the contact, before chalcopyrite is found more distal and at depth. Minor inflections are observed along the margins of the HEMQ, where strongly localised high-grade concentrations are found in 'antiformal' geometries. However, despite this apparent intimate relationship, it should be noted that not all margins of the HEMQ are mineralised, and that a significant volume of high-grade mineralisation is observed distal to its contact (Ehrig and Clark, 2020).
iv). Localised occurrences of paragonite-illite-quartz ±zunyite ±diaspore ±kaolinite advanced argillic alteration remnants that are partially to wholly replaced by hematite.

Distal to the main alteration/mineralised system of the Olympic Dam Breccia Complex, relatively minor post-magmatic alteration of the Roxby Downs Granite takes the form of an evolution from cryptoperthite* to patch perthite*. Patch perthite was subsequently overprinted by high porosity, near end-member albite and K feldspar, whereas plagioclase is partially replaced by albite-muscovite ±Ba-rich K feldspar, both of which are isochemical processes requiring no addition of K and Na from an external source (Kontonikas-Charos et al., 2017). Inward within this zone, biotite and amphibole are progressively altered to chlorite, until no biotite remains at the margin of the breccia complex, whilst magnetite is converted to hematite. Muscovite alteration of magmatic and hydrothermal feldspars and chlorite persists into the margins of the Olympic Dam Breccia Complex, whilst plagioclase is completely altered to muscovite plus albite, which is, in turn, then altered to muscovite (Ehrig et al., 2017).
*NOTE: Cryptoperthite is a fine-grained, submicroscopic feldspar variety that has the structure of perthite, and consists of an intergrowth/interlamination of potassic and sodic feldspar (albite and K feldspar), detectable only by x-rays or an electron microscope. Patch perthite is an intimate intergrowth of sodic and potassic feldspar resulting from sub-solidus exsolution.

The stages that followed the early Fe²⁺ phase of the footwall, are composed of the following hydrolytic alteration that overprints and largely obliterates early alkali-calcic and potassic alteration. These processes produce assemblages that include hematite, muscovite, phengite, chlorite, quartz, siderite, fluorite, barite and remnants of kaolinite, pyrophyllite and topaz (Ehrig et al., 2011; 2017, Ciobanu et al., 2019, Verdugo-Ihl et al., 2020; Ehrig and Clark, 2020, Ehrig 2022):
• Hematite alteration - Hematite has a history of multiple pseudomorphic replacement of pre-existing lithologies and minerals, accompanied by corrosion, overgrowth and recrystallisation. It replaces and pseudomorphs magnetite, siderite, chlorite; sometimes sulphides, feldspars, fluorite and barite; and rarely as infill in the breccia matrix. In the more heavily altered intervals it occurs as both oscillatory and 'sector-zoned' hematite, with haloes of overprinting porous hematite with corrosive margins between the two. The oscillatory and zoned hematite may be displaced across fractures that are commonly filled with the porous variety. The porous hematite can contain sub-micron inclusions of U and REE minerals that coarsen to µm-scale inclusions as the porosity increases (Ciobanu et al., 2013). Hematite is also precipitated in vugs. Hematite alteration of magnetite is represented by: Fe3O4 + 2H⁺ → Fe2O3 + Fe²⁺ + H2O. This conversion is stimulated by a decrease in temperature in higher, cooler, later parts of the system as it degrades from magnetite-stable conditions at temperatures in the range 450 to 350°C at depth, to hematite stability at <350°C (e.g., Corriveau et al., 2022). However, much of the hematite alteration is the result of dissolution of the other replaced minerals and/or substitution by hematite from a hydrothermal solution.
The two broad, peripheral zones of the Olympic Dam Breccia Complex comprise the outer weakly altered brecciated-granite to granite-breccia (GRN) and inner granite- to hematite-rich breccia (GRBx) with <5 and 5 to 25 wt.% Fe respectively, and hence, relatively minor hematite alteration, as described previously in the Geology section (see Fig. 1). The inner granite-rich breccia, passes inwards into narrower, more discrete zones of hematite-rich breccias, defined by the >25 wt.% Fe contour, where hematite becomes the most abundant component. These hematite-rich breccias are also predominantly derived from Roxby Downs Granite, but reflect a greater intensity of brecciation and a higher proportion of hydrothermal constituents, principally hematite, relative to the granite-rich clast breccias. Hematite-rich breccias are the major host to sulphide mineralisation. Hematite may be red-brown to black. Increased hematite alteration of the Roxby Downs Granite coincides with a marked depletion of free silica, from 30 to 35 wt.% Si (65 to 75% SiO2) in the fresh granite, to <2.5% wt.% Si (<5% wt.% SiO2) in hematite breccia with 60 wt.% Fe (Ehrig et al., 2017). This represents a substantial volume of silica that has been leached from the granite, some of which may have been deposited as late veinlets cutting the HEMQ hematite-quartz breccias and as zones of quartz veining on the margin of the deposit.
• Chlorite alteration - Mg-Fe chlorite has replaced magmatic biotite and amphibole in the Roxby Downs Granite, and in mafic-ultramafic clasts in hematite-breccias, as well as the mafic silicates in the abundant mafic-ultramafic dykes, lavas and sills. The intensity of alteration increases proximal to contacts with mafic and ultramafic dykes and intrusions, but grades into/is replaced by, intense sericite altered breccias distal from the contacts with pre- to syn-brecciation dykes. Chlorite alteration appears to increase with depth in granite-derived breccias, typically pseudomorphing feldspars. Ehrig et al. (2017) show the main, relatively narrow sheet of more intense chlorite immediately overlying the siderite-K feldspar zone and passing up into, and overlapping muscovite. Muscovite alteration typically overprints and replaces chlorite. Alteration of biotite to chlorite by an acidic fluid is represented by the equation: 2K(Mg,Fe)3AlSi3O10(OH)2 + 4H⁺ → (Mg,Fe)5Al(Si3Al)O10(OH)8 + 3SiO2 + (Mg,Fe)²⁺ + 2K⁺, releasing silica and potassium (Ehrig, 2022).
• Sericite/muscovite alteration (which with other white micas are collectively referred to as muscovite) - muscovite typically replaces feldspars and chlorite, and occasionally occurs as breccia matrix infill. It is widespread throughout the Olympic Dam Breccia Complex, occurring as an alteration product of igneous sodic plagioclase (albite and oligoclase), igneous orthoclase, and hydrothermal chlorite, and is, in turn, replaced by hematite, culminating in the hematite-quartz-barite breccia. Incipient muscovite alteration of sodic plagioclase occurs in relatively unaltered Roxby Downs Granite, whilst it completely replaces plagioclase when the accompanying Fe concentration reaches 5 wt.%. Similarly, the orthoclase content steadily decreases as the Fe concentration increases, until it essentially disappears to <1 wt.% when Fe becomes >35 wt.%. Clasts of feldspar-phyric felsic lavas in hematite breccias are variably altered to sericite and hematite. Ehrig et al. (2017) show the zone of strongest muscovite alteration immediately underlying and flanking the intense Fe alteration of the HEMQ Breccia, and directly overlying the chlorite zone. However, moderate muscovite alteration continues downward and outward, overlapping both chlorite and the siderite-K feldspar zones. Alteration of alkali feldspar to muscovite by an acid fluid is represented by the equation: 3KAlSi3O8 + 2H⁺ → KAl2(AlSi3O10)(OH)2 + 6SiO2 + 2K⁺, again releasing silica and potassium (Ehrig, 2022).
• Intense sericitic (muscovite), grading to advanced argillic alteration occurs mainly at shallow depths to the ENE of the central hematite-quartz-barite breccia zone. Granite-rich breccias are bleached with >35 wt.% sericite accompanied by 5 to 10 wt.% Fe, with <0.3 wt.% Cu, but may contain >0.3 g/t Au. Relict orthoclase occurs, along with ~1.5 wt.% hydrothermal albite, with rare zunyite, kaolinite and diaspore. This alteration is commonly accompanied by strongly altered mafic-ultramafic dykes and intrusions. It also occurs as isolated remnants in hematite-altered breccias in the northwestern extension of the deposit and near the edges of the hematite-quartz-barite breccia (Ehrig et al., 2012). Alteration of muscovite to kaolinite (or pyrophyllite and andalusite, depending on the pH) is represented by the equation: 2KAl2(AlSi3O10)(OH)2 +2H⁺ + 3H2O → 3Al2Si2O5(OH)4 + 2K⁺. The subsequent dissolution of kaolinite comprises: Al2Si2O5(OH)4 + 6H⁺ → 2SiO2 + 2Al³⁺ + 5H2O, again releasing silica to the acidic hydrothermal fluid (Ehrig, 2022).
• Hydrothermal quartz was released during alteration of feldspars, muscovite and chlorite (as indicated by the formula above), and occurs locally in some locations as 1 to 5 mm thick microveinlets with apatite, magnetite and/or siderite in significantly altered, but largely un-brecciated granite, and as anhedral to subhedral crystals in breccias. It is uncertain whether these account for all of the silica formed during hydrolytic alteration, and most would appear to have been removed in solution by hydrothermal fluids. It is also found as 20 to 50 m thick quartz veins in altered Roxby Downs Granite in the extreme SE of the Olympic Dam Breccia Complex, where it forms a NNW-SSE elongated, >800 x 200 m zone within altered, weakly brecciated granite. These quartz veins are not geochemically anomalous in any trace metals. Timing of the quartz veining has been difficult to establish. Silicification also occurs as irregular zones along the edges of the hematite-quartz-barite HEMQ breccias (Ehrig et al., 2012).
• Siderite, fluorite and barite alteration occurs as veins and as breccia matrix infill, and is zoned across the deposit. Barite tends to be best developed in the shallow core of the Olympic Dam Breccia Complex in the HMEMQ Breccia, grading out and down to overlapping fluorite which passes outward into siderite. As detailed above, siderite typically occurs at depth and on the edges of the deposit, accompanied by and replacing magnetite, and is less susceptible to replacement by hematite than is magnetite. The strongest fluorite corresponds closely to the highest Cu ores. Pyrite and chalcopyrite can also accompany siderite, along with fluorite, although the latter is more strongly associated with hematite alteration and chalcopyrite, bornite and chalcocite. Barite is closely associated with hematite in hematite-quartz-barite breccias which lack sulphides, and the rocks immediately surrounding those breccias. Elsewhere, barite occurs with fluorite, chalcopyrite, bornite and chalcocite. Siderite and barite do not usually occur together, except in late stage barite-fluorite-siderite veins that cut the Olympic Dam Breccia Complex (Ehrig et al., 2017).
• The bedded clastic facies have, in places, been subjected to intense texturally destructive hematite alteration that overprints the contacts with neighbouring facies and altered Roxby Downs Granite. This alteration includes sericite and chlorite that is replaced by hematite. However, overall the 'bedded clastic facies' have not been brecciated, mineralised or altered to the same degree as the Roxby Downs Granite. Again, apart from specific areas, the green mudstones and sandstone (KASH facies) is little altered, without an IOCG signature, while the red mudstone and sandstone (KHEMQ facies) has an IOCG chemical signature and is hematite rich, possibly due to pervasive alteration of porous sandstone and its mudstone wallrocks.
• Carbonate minerals - the deposit contains a volumetrically substantial and mineralogically diverse component of carbonate minerals which are always associated with ore sulphides and uraninite. Siderite and siderite-rhodochrosite-magnesite in solid solution are the dominant carbonate types, whilst calcite, dolomite-ankerite solid solution and REE-fluorocarbonates are locally abundant. Isolated carbonate grains are typically compositionally zoned (both simple or oscillatory), whilst replacement textures, including mutual replacement of carbonates by other carbonates and hematite are common (Apukhtina et al., 2020). Dating of Ca-Fe-Mg-Mn-carbonate gangue minerals which are associated with ore minerals reflects virtually every stage of mineralisation in the prolonged, multi-stage history of brecciation and mineralisation from 1.59 to 0.4 Ga. Sm-Nd and Pb-Pb isotope dating of calcite veins in ~1.59 Ga IOCG ore indicate 1.59 to 1.55 Ga deposition ages, whilst the structurally youngest carbonates yield ~0.50 Ga Sm-Nd ages. The latter are unbrecciated fluorite-barite veins and carbonate matrix in polymict conglomerate above the breccia complex, the inferred carbonate ages of which are broadly consistent with radiometric dates for other hydrothermal minerals e.g., hematite, uraninite, apatite, fluorite associated with these veins (Maas et al., 2022).
Carbonates are found in all lithologies at Olympic Dam, occurring in the following associations:
i). Pink to white, un-brecciated, coarse-grained calcite veins in weakly, in situ brecciated Roxby Downs Granite and rhyolite. These veins crosscut texturally early, ~1590 to 1570 Ma magnetite-pyrite-apatite-quartz-uraninite-dominant assemblages and locally include barite, fluorite, quartz, hematite, anhydrite and synchysite, with ~1% sulphides, mainly galena, chalcopyrite and pyrite (Apukhtina et al., 2020). These veins yielded ages of 1546 ±33 Ma (Sm-Nd isochron; Maas et al., 2022) for calcite; and 1585 ±23 Ma (Pb-Pb isochron age; Maas et al., 2022) for calcite and anhydrite, which are broadly consistent with the age of the host granite, and associated apatite (1583 ±7 Ma) and uraninite (1594 ±6 Ma) from the U-rich assemblage they cut (U-Pb ages; Jagodzinski, 2014; Apukhtina et al., 2017).
ii). Carbonates in coarse- to fine-grained, strongly brecciated granite, accounting for ~88% of all carbonates. Siderite is by far the most abundant carbonate, occurring in breccia cement, breccia clasts, breccia-hosted former pores and as carbonate-bearing ooids, in order of decreasing abundance. Multiple generations are evident, varying from almost pure FeCO3 to Mn-Mg-rich siderite which have replaced and overgrown each other, with the end members forming oscillatory zoned carbonates. Siderite-bearing breccia cement carries abundant relicts of earlier minerals, such as magnetite and earlier siderite, that have been partially converted to hematite with local µm U-Mo-W-Sn inclusions, and pyrite with apatite inclusions that has been replaced by chalcopyrite. Siderite is both cogenetic with, and has replaced hematite. Late-stage voids carry barite, quartz, fluorite and chalcopyrite, and rarely galena, sphalerite, Co-Fe-As sulphides including tennantite-tetrahedrite, bornite and chalcocite. They also contain Bi-Ag-Au-Pb-Te-Se alloys, that are commonly found as inclusions in chalcopyrite, as well as U-Th-REE bearing minerals such as monazite, florencite and REE-fluorocarbonates with commonly associated sericite and chlorite. Siderite-rich breccia clasts are up to several cm across, and are typically associated with carbonate vein fragments. They are brown, have strongly variable 'within-clast' grain-sizes, and typically occur within a siderite dominated cement. Siderite within clasts is locally cut by Mg-Mn-rich siderite veinlets and has replaced, or has itself has been replaced by, hematite (Apukhtina et al., 2020). The samples dated by Maas et al. (2022), excluding those suspected of contamination, yielded ages in the range 1.60 to 1.55 Ga (Sm-Nd isotope).
iii). Carbonate veins in 'bedded clastic facies', typically ~1 cm thick, mostly dolomite-ankerite, with µm-scale calcite and Fe-rich carbonate inclusions, in both the KASH and KHEMQ facies. Other gangue minerals in the veins, which are typically internally zoned, include barite, fluorite, quartz, hematite and anhydrite. Anhydrite is restricted to the green sandstone and mudstone facies (KASH). Sulphide minerals within the carbonate veins mirror those of their host bedded facies. Carbonate veins in the sandstone and red mudstone (KHEMQ) also include some tennantite-tetrahedrite (Apukhtina et al., 2020). Sm-Nd isotope data for hydrothermal carbonates yields ages in two fields, the first in the main array between 1.59 and 1.55 Ma, consistent with the age of the host clastics, and a second field at ~ 0.49 Ga (Maas et al., 2022).
iv). Carbonate in mafic and ultramafic igneous rocks, occurring as ~1 cm thick veins of either a). calcite and dolomite-ankerite accompanied by chlorite, sericite and hematite, or b). barite-fluorite. Both vein sub-types also contain quartz, anhydrite, hematite, Ti-oxide and diverse sulphides (Apukhtina et al., 2020). Dolomite-ankerite from picrites were dated at ~1593 Ma; siderite, calcite and dolomite-ankerite from basalts also gave ages of ~1593 Ma; calcite and dolomite-ankerite from Gairdner Dolerite Swarm dykes returned ages of ~825 Ma, while dolomite-ankerite from another generation of basaltic rocks were dated at ~500 Ma (Rb-Sr and Sm-Nd isotopes; Maas et al., 2022).
v). Un-brecciated, massive barite-fluorite dominated veins with minor carbonate which crosscut the Olympic Dam Breccia Complex along NE and east-west arrays. The veins vary from a few mm to ~1 m in thickness, and can be traced for >800 m vertically. They also carry minor Ca-Mg-Fe-Mn- and REE-F-rich carbonates, quartz, botryoidal hematite, chalcopyrite, pyrite, as well as traces of bornite, chalcocite, tennantite-tetrahedrite, safflorite, cobaltite, löllingite, sphalerite, galena, alloclasite, native gold, diverse Bi-Ag-Pb-Se-Te bearing alloys and U-REE minerals, including uraninite, coffinite, brannerite, florencite and REE-fluorocarbonates (Apukhtina et al., 2020). Dolomite-ankerite and siderite from this trend are dated at ~0.5 Ga (Sm-Nd isotopic composition) supported by cogenetic fluorite also with a 0.5 Ga age (Sm-Nd age; Maas et al., 2022).
vi). Intervals largely composed of beige to brown siderite up to several tens of metres thick replacing lithologies within the Masher's Fault zone. This siderite has a wide range of textures, including laminated to drusy, cockade and crustiform. It is un-brecciated and replaces clasts up to several cm across of hematite-dominant 'bedded clastic facies'. In some cases, replacement and concentric overgrowths by fine- to coarse-grained siderite has obliterated all protolith textures. These carbonates are overwhelmingly pure FeCO3, locally replaced by Mg-Mn-rich siderite. The laminated siderites, which are the dominant form, are interpreted to be altered Fe-rich sandstones, overprinted by hydrothermal siderite (Apukhtina et al., 2020). Samples of this association have very coherent Sm-Nd isotope systematics within the Main Array, implying 1.59 to 1.55 Ga ages, broadly supported by the 1598 ±170 Ma Sm-Nd age for 12 data points dated by Maas et al. (2022).
vii). Coarse- to fine-grained carbonate matrix in alternating layers of conglomerate-breccia and sandstone. The beds are immediately above the main hematite-breccia hosted ore deposit, separated from it by the Neoproterozoic unconformity, and are overlain, in turn, by dolomitic-like facies similar to the Neoproterozoic Ediacaran Nuccaleena Formation. The conglomerate-breccia is composed of polymictic, spherical to oval, to slightly angular clasts up to several cm across. Strong hematite alteration and clast-in-clast textures mask identification of the original protolith of most clasts, although Roxby Downs Granite and Gawler Range Volcanics clasts are evident. Other clasts appear to be glacial dropstones. All are set in a sedimentary carbonate matrix. The sandtones are cross-bedded, fine-grained and well-sorted, composed of rounded, ~3 mm diameter quartz-dominated grains, with local hematite-rich layers. The carbonate matrix/cement in both lithologies is predominantly strongly zoned dolomite-ankerite grains (Fe-poor cores → Fe-rich rims) with associated, intergrown barite and celestite. Pyrite, chalcopyrite, bornite and minor altaite are locally disseminated throughout both the conglomerate clasts and cement, whilst sulphide-carbonate veins are commonly restricted to individual clasts. Immediately above the unconformity, the dolomitic facies are thinly bedded, fine-grained (~50 µm) with quartz and possibly granite derived accessories, including apatite, zircon and Ti-oxides. There is evidence that dolomite in this dolomitic facies has undergone post depositional Fe alteration, e.g., ankerite rims on dolomite crystals and ankerite veins crosscutting dolomite, as well as crosscutting barite-carbonate veins (Apukhtina et al., 2020). The Sm-Nd isotope signatures of these carbonates, with their characteristic high and variable Sm/Nd, clearly yield an Sm-Nd age of ~0.5 Ga (Maas et al., 2022). If the ~1590 Ma deposit has been upgraded by an ~0.5 Ga unconformity-style uranium mineralisation event, this alteration association would represent the fluid interface.
The ages of these carbonates, all of which have varying degrees of associated 'ore sulphides' and uraninite, are interpreted to reflect three periods in which mineralisation was deposited or re-deposited, namely ~1.60 to 1.55 Ga, ~0.82 and ~0.5 Ga. The dates are consistent with ages measured for other minerals, both gangue and ore, providing further evidence for a prolonged multi-stage history of ore formation and reinforcement of grade and tonnage through repeated overprints (Maas et al., 2022).

MINERALISATION

There is a strong enrichment of a broad range of elements that have a direct positive correlation with Fe up to ~45 wt.% Fe (or ~75 wt.% hematite). These elements include Cu, U, Au, Ag, F, S, C, As, Ba, Bi, Cd, Co, Cr, Fe, In, Mo, Nb, Ni, P, Pb, Sb, Se, Sn, Sr, Te, V, W, Y, Zn and REE. All of these are regarded as being primarily of hydrothermal origin. In particular, there is a strong spatial correlation between Cu, U, Au and Ag. Above ~75 wt.% hematite, the concentrations of Cu, U3O8, and Au sharply decrease (Ehrig, McPhie and Kamenetsky, 2012), whilst below ~30 wt.% hematite, there is also a marked decrease. These trends are much more extreme for chalcocite, decreasing through bornite to chalcopyrite, with little correlation between Fe and Cu grade in pyrite(-chalcopyrite) mineralisation. In contrast, elements interpreted to be affiliated with granite, such as Al, Be, Ca, Hf, K, Li, Mg, Mn, Na, Rb, Si, Th, Ti and Zr are negatively correlated with Fe, with the degree of depletion increasing proportional to the Fe concentration (Ehrig, McPhie and Kamenetsky, 2012).
The hematite content is proportional to the increase in permeability generated by brecciation towards the spatial centre of the deposit (Reeve et al., 1990), with sulphides being precipitated within the breccia matrix as the volume of matrix increased (Ehrig, McPhie and Kamenetsky, 2012). Clark and Ehrig (2019) conclude therefore, that between ~30 and 75 wt.% hematite, the permeability is optimal for maximum sulphide precipitation, if the hydrothermal fluid-rock interaction is sufficiently high. Above 75 wt.%, the permeability becomes significantly occluded, reducing fluid flow and the ability to precipitate economic concentrations of sulphides.
Ca and Mn were initially depleted through amphibole and biotite being altered to chlorite, whilst the early sericite alteration of sodic plagioclase depleted Na. This was followed by enrichment of F and Mn via the precipitation of fluorite and Mn-bearing siderite when Cu mineralisation reached ≥1000 ppm. In Cu-barren and weakly to moderately mineralised levels of ≤3000 ppm Cu, the Olympic Dam deposit is enriched in LREE (light rare earth elements), whereas the MREE (medium rare earth elements) and HREE (heavy rare earth elements) are depleted. The MREE and HREE become enriched in mineralisation of ≥3000 ppm Cu. The REE pattern for samples with >50% Fe and ≥3000 ppm Cu are depleted relative to samples with <50% Fe (Ehrig et al., 2012).
Dmitrijeva et al. (2019) studied the lithogeochemical zoning of the Olympic Dam deposit, and found the signature of the deposit was composed of two geochemical associations, which exhibit distinct, but spatially overlapping distributions: i). The first group, Cu-U3O8-Se-S, which comprises the main economic and closely associated elements, has a concentric zonation, outward and downward from a Cu-barren, high-Fe, hematite-quartz-barite-bearing breccia (HEMQ) in the upper central core of the deposit; this signature is also found in many of the altered mafic-ultramafic dykes. ii). The second group, Au-W-Mo-Sb-As, forms a column with a vertical extent of >1800 m, still open at depth, and is located in the southeastern lobe of the deposit where it spatially coincides with >5% Fe hematite and volcanic breccias within the Olympic Dam Breccia Complex. This second signature is especially evident at depths below 1200 m, as well as in generally shallow, hematite-quartz-barite bearing breccias (HEMQ) and hematite-quartz sediments (KHEMQ) of the 'bedded clastic facies', but not the chlorite-bearing more mafic facies sediments (Dmitrijeva et al., 2019). The Au-W-Mo-As-Sb signature is overwhelmingly related to the occurrence of compositionally-zoned hematite, which hosts most of the W, Mo and Sn throughout the deposit, as well as a significant part of the U (Verdugo-Ihl et al., 2017), and to sulphides hosting Sb and As, which include pyrite (As); tetrahedrite-tennantite (Sb-As); and cobaltite (As) (Ehrig et al., 2012). Gold is mostly native, alloyed with Ag (and rarely Cu), or as Au(±Ag)-tellurides, occurring as inclusions in sulphides and hematite. This vertical mineralised column is considered likely to be related to the intersection between the major subvertical WSW-ENE and NNW-SSE striking fault systems which acted as a hydrothermal-fluid conduit. However, the concentration and zoning of the first group signature, while being controlled by the same mineralisation/alteration corridor, appears to be spatially independent of the Au-W-Mo-As-Sb association, i.e., two separate mineral associations taking advantage of the same conduit.
Ehrig et al. (2017) regard the principal controls of Cu grade within the deposit, and by inference, U, Au and Ag, in decreasing order of significance, to be: i). sulphide species and abundance; ii). intensity and type of iron oxide alteration, i.e., hematite or magnetite; iii). structure and geometry; and iv). protolith lithology.
Ehrig and Clarke (2020) demonstrated by plotting the very high grade intervals (i.e., >5% Cu, >2.5 g/t Au), that such high grades closely follow faults, particularly fault intersections, and are closely associated within and around the very reactive ~1590 Ma mafic to ultramafic dykes, the distribution of which are controlled by the same faults. The very high grade Cu is particularly evident in the core of the NW neck of the resource, whilst gold is more heavily concentrated in the core of the deposit.
The overall paragenetic sequence of sulphide precipitation within the deposit is sphalerite → galena → pyrite → chalcopyrite → bornite → chalcocite, although sphalerite and galena are only found in trace amounts. However, locally, sphalerite contents may be up to 2 wt.% in distal breccia (Ehrig et al., 2012), and late-stage veinlets of galena occur within the HEMQ hematite-quartz-barite breccia of the barren core (Corriveau et al., 2022).

High grade mineralised hematite breccia

The occurrence of each of the main economic metals can be summarised as follows:
• Copper - The dominant copper sulphide minerals within the deposit are chalcocite/digenite/djurleite (collectively referred to as chalcocite), bornite, chalcopyrite and pyrite, which occur in the relative proportion of 25, 50, 18, 7%, respectively, with lesser sphalerite, galena, molybdenite, tennantite-tetrahedrite and traces of native copper. These sulphides predominantly occur as disseminated grains in both the breccia matrix and clasts, with grain sizes that vary from <20 µm to several millimetres, averaging ~100 µm. Sulphides may also, but rarely, occur as veins in granite-dominated breccias.
Tennanite-tetrahedrite and enargite occur in relatively small amounts and appear to be associated with ores in the deep mineralised zone, above and surrounding mineralisation below –950 mRL. Covellite, native copper and tenorite/cuprite are very rare. Most sulphides do not display textures indicative of supergene formation, suggesting they are hypogene in origin. However, covellite, which occurs along the edges of and in cracks within bornite grains, at or near the unconformity with the overlying Neoproterozoic cover sequence, does suggest it is related to a supergene influence. Never-the-less, if a significant supergene blanket was developed, it has since been removed by Meso- to early Neoproterozoic erosion. Native copper and tenorite/cuprite are locally found in association with chalcocite, but usually not near the overlying unconformity. Pyrite, chalcopyrite, bornite and chalcocite may be intergrown with the major gangue minerals of hematite, quartz, sericite, orthoclase, chlorite, barite, fluorite and siderite, whilst bornite and chalcocite are only rarely associated with siderite.
There is a distinct upwards and inwards zonation of sulphide/copper species within the deposit, from: pyrite → chalcopyrite-pyrite → chalcopyrite as exsolution lamellae in brown bornite → purple bornite-chalcocite symplectites → digenite lamellae in chalcocite → occasional native copper (Reeve et al., 1990; Ehrig, McPhie and Kamenetsky, 2012; Ciobanu, Cook and Ehrig, 2017).
Two generation of pyrite have been recognised, namely: i). coarse-grained euhedral crystals that are commonly associated with magnetite, with or without chalcopyrite, although the chalcopyrite is not seen to replace the pyrite; and ii). a more common anhedral to subrounded species that is partially replaced by chalcopyrite. Pyrite replacement of chalcopyrite has not been observed. Very rare small pyrite grains have been found in massive bornite.
Chalcopyrite may occur as i). isolated grains; ii). co-precipitations with bornite: iii). grains being replaced by bornite; or iv). rare replacement of bornite.
Bornite occurs as i). isolated grains; ii). grains being replaced by chalcocite, or iii). most commonly, as myrmekitic bornite-chalcocite intergrowths.
Chalcocite occurs as isolated grains and rarely replaces bornite.
Chalcopyrite-bornite pairs are not as common as chalcopyrite-pyrite or bornite-chalcocite. The pairs chalcopyrite-bornite, bornite-chalcocite and chalcocite-digenite may represent the progressive addition of Cu at temperatures of >300°C, and form an upward and inward gradient of increasing Cu:Fe and decreasing S:Cu+Fe ratios (Ciobanu, Cook and Ehrig, 2017). However, as the transition from chalcopyrite to bornite-chalcocite in the Olympic Dam deposit invariably coincides with a general 2 to 4 times increase in Cu grade, the actual amount of sulphur present remains comparable throughout (e.g., 1% Cu as chalcopyrite = ~1% S, whilst 2.5% Cu as bornite = ~1% S; and 4% Cu as chalcocite = ~1% S). It should be noted most o the higher tenor chalcocite and bornite were apparently formed at lower temperatures of 250 to 150°C.
Other base metals are found on the margins of the Cu-U-Au-Ag deposit, where chalcopyrite-sphalerite-galena can be co-genetic or sphalerite-galena may be later than chalcopyrite. A galena-chalcopyrite association is regarded as cogenetic with pyrite, although the galena is generally non-radiogenic, implying it formed early, prior to the system becoming enriched in radiogenic lead.
Ehrig, McPhie and Kamenetsky (2012) describe an outer, distal and deep or paragenetically early Mo-W-Sn-As-Sb enrichment at geochemically anomalous levels which may be overprinted by Cu-Pb-Zn, in the south-eastern lobe of the deposit defining a further sulphide zonation. This zone is postulated to grade inwards and upward → (Cu-Pb-Zn) → and the barren (hematite-quartz-barite) breccia upper core to the deposit (Corriveau et al., 2022). Deep drilling has intersected this Mo-W-Sn-As-Sb signature overprinted by grades of between >0.5 to 4% Cu at depths of from ~900 to 1800 m below surface (Fig. 4, section G-H). The Mo-W-Sn-As-Sb signature is geochemically distinct and in the deep section is bounded by faults. It is most likely related to the Au-W-Mo-Sb-As signature of Dmitrijeva et al. (2019). The host hematite breccia contains recognisable clasts of KHEMQ, red sandstone-mudstone of the 'bedded clastic facies', but also appears to overlap the VBx volcanic breccia. It is also intruded by ~1950 Ma mafic to ultramafic dykes. The adjacent fault block, between the Woodall and Woodall Fault Junior is characterised by K feldspar alteration and contains veins of quartz-tourmaline, quartz-carbonate and quartz. This high grade zone is interpreted to lie within a fault slab of 'bedded clastics facies' that has been tilted and collapsed into the Olympic Dam Breccia Complex during the life of that complex's formation (After Ehrig, 2022 SAEMIC conference). As such it may reflect an initial interaction with hydrothermal fluids above the main breccia complex prior to collapse, and subsequent alteration and mineralisation within the breccia complex. It appears to lie within the transition from the chalcocite-bornite to bornite-chalcopyrite zones.

High grade mineralised hematite breccia

• Uranium is predominantly present as fine-grained particles of uraninite [(U,Pb,Ca,Y,REE)O2], coffinite [U(SiO4)1-x(OH)4x] and brannerite [(U,Ca,Y,Ce)(Ti,Fe)2O6], disseminated throughout, interstitial to the sulphides and gangue minerals, and accounting for >80% of the uranium in the deposit (Ehrig, et al., 2021). The grain size of the uranium minerals ranges from <0.5 to >1000 µm, averaging ~20 µm, with uraninite-coffinite and coffinite-brannerite commonly complexly intergrown. Coffinite and brannerite are interpreted to have formed as a result of the alteration of uraninite. The uraninite occurs as fine <30 µm euhedral grains that are primary, and as anhedral to subhedral to round grains, which are <30 to ~100 µm in diameter and are chemically and texturally distinct from the euhedral form. The anhedral to subhedral/rounded grains, which are prominent in high-grade ore zones, form: i). larger massive aggregates; and rarely, ii). microveinlets that can be as much as 1 mm wide (Macmillan et al., 2016, 2017).
Another ~15% of the uranium occurs as submicron-sized particles and lattice substitutions in hematite, along with elevated Mo-W-Sn (Ciobanu et al., 2014), whilst <5% occur in sulphides, including pyrite, chalcopyrite, bornite and chalcocite, as well as in uranothorite, thorianite, the REE minerals bastnäsite, florencite and synchysite, and in xenotime, monazite, zircon and the crandallite group. Uraninite has a slightly stronger association with hematite, whilst coffinite and brannerite are a little more commonly associated with non-hematite gangue (Ehrig, et al., 2012, 2017).
Deposit-wide, the average relative proportions of brannerite, coffinite and uraninite are 31, 56 and 13% respectively, although the latter increases to >40% at high Fe concentrations. The uranium minerals have no statistically significant preferential association with pyrite, chalcopyrite, bornite or chalcocite. Unlike the Cu ±Fe sulphides, the uranium minerals do not form systematically zoned patterns across the deposit (Ehrig, et al., 2017). The uranium mineralisation was deposited and remobilised during multiple cycles of dissolution and re-precipitation of these uranium ore minerals, and the composition of each of these minerals is complex and very variable (Ehrig and Clark, 2020)
The high abundance of uranium in the Olympic Dam deposit allows U-Pb dating of ore minerals. As detailed above, two types of uranium mineralisation are evident, namely i). fine euhedral primary grains of uraninite, and ii). anhedral to subhedral/rounded grains of uraninite, coffinite and brannerite. U-Pb dating of the fine-grained euhedral uraninite yields ages of ~1.59 Ga (Macmillan et al., 2016; Apukhtina et al., 2017). These euhedral uraninite grains contain high total rare earth element (REE) levels, have relatively unfractionated REE patterns with low Ce/Lu, pronounced La and Eu depletion and low Y/Ho ratios. In contrast, the texturally distinct anhedral to subhedral/round uraninite grains have variably preserved U-Pb ages of ~0.5 Ga, and are characterised by lower total REE levels, and by REE patterns with very low La/Sm ratios, pronounced peaks of Sm, and an absence of Eu anomalies (Ehrig et al., 2021). Apart from these younger age dates, there is a pronounced deficit of Pb present compared to that expected in a U deposit formed at 1.59 Ga. This is best accounted for by at least part of the U having been added as late as 1 billion years subsequently. Uranium mineralisation of this younger generation is locally found with partially preserved, yet significantly altered remnants of the older ~1.59 Ga euhedral uraninite (Macmillan et al., 2016), and dominates the highest U ore grades in the deposit, also consistent with the younger U-Pb ages. Ehrig et al. (2021) note that the REE patterns in this younger uraninite generation resemble REE signatures typical of low-temperature uraninite in the large, high-grade Proterozoic unconformity-related U deposits of Canada and northern Australia (Fryer and Taylor, 1987; Frimmel et al., 2014).
Additional evidence for a major Late Neoproterozoic to Cambrian U mineralisation event is provided by the Pb isotope composition of thousands of hydrothermal chalcopyrite, bornite and chalcocite grains analysed at Olympic Dam. These data define a ²⁰⁷Pb/²⁰⁶Pb ratio in the range 0.07 to 0.06. U concentrations in these sulphides, including those with highly radiogenic Pb, vary widely, but the vast majority have low U/Pb ratios, interpreted to imply the radiogenic Pb did not evolve within the low-U sulphide carrier minerals, but was inherited from associated U minerals. The timing of the production and release of highly radiogenic Pb from U minerals, and its subsequent capture by Cu sulphides, can be calculated to have occurred between 0.9 and 0.6 Ga, if Pb release occurred in the recent geological past. If Pb release occurred earlier, the parental U minerals cannot be older than ~0.7 Ga. Ehrig et al. (2021) conclude the 'remarkably' homogeneous and low ²⁰⁷Pb/²⁰⁶Pb in the Cu sulphide minerals implies a major period of U mineral formation at 0.7 to 0.5 Ga, broadly consistent with the ~0.5 Ga ages of non-euhedral uraninite. The presence of this young radiogenic Pb in the Cu sulphides is also interpreted to imply widespread concomitant modification of precursor, and possibly growth of new, sulphide minerals.
In addition to these two generations of uranium mineralisation, age dating of uranium minerals also suggests the possible addition of minor U during the ~1200 to 1160 Ma Musgravian Orogeny (Ehrig, 2019)
Ehrig et al. (2021) note that the Olympic Dam Breccia Complex is truncated by the Neoproterozoic Marinoan glaciation event during the Cryogenian, and unconformably overlain by the cover sequence deposited under periglacial conditions during a period that coincided with a rise in atmospheric oxygen (Lyons et al., 2014). As such, they suggest, the younger, 0.7 to 0.5 Ga U mineralisation may have been related to circulation of oxidised basinal fluids from above the unconformity which enhanced the solubility and mobility of the pre-existing 1.59 Ga uranium mineralisation.
Age dating of uranium minerals also suggests the possible addition of minor U during the ~1200 to 1160 Ma Musgravian Orogeny (Ehrig, 2019).
• Gold occurs occurs as Au-Ag-Cu-Fe electrum, calaverite, petzite and tetra-auricupride [AuCu] as i). inclusions within, or attached to, sulphide minerals; ii). inclusions in quartz, hematite and barite in gold-dominant, sulphide free zones on the leading edges of barren hematite-quartz-barite HEMQ breccias, juxtaposed against Cu sulphide mineralisation in the adjacent hematite breccias; and iii). inclusions in sericite and/or quartz in gold-dominant zones in intensely muscovite and paragonite altered granite breccias (Ehrig, et al., 2012, 2017).
Gold particles are typically ~1 to 20 µm across, although bonanza veins and rare large nuggets are found in hematite-rich gold-only zones (Ehrig, McPhie and Kamenetsky, 2012). Two distinct electrum populations have been recognised: i). >95% Au with <5% Ag+Cu where associated with sulphides, and ii). <90% Au with ~10% Ag+Cu+Fe where associated with sericite, quartz, hematite or barite, in the absence of sulphides (Ehrig, et al., 2012, 2017).
• Silver is predominantly found as >10 µm hessite micro-inclusions in bornite, but also occurs in electrum, petzite, acanthite, allargentum and rarely as native silver. Bornite, chalcocite and tennantite-tetrahedrite may also contain sub-micron inclusions of native silver.
• Rare Earth Elements - the principal REE and their grades at Olympic Dam are ~0.17 wt.% La and 0.25 wt.% Ce, contained in bastnäsite and florencite, with lesser synchysite, monazite, xenotime, zircon and crandallite. These minerals occur as grains disseminated between both gangue and sulphide minerals, with an average grain size of 35 µm. REE also occur in minor to trace concentrations in uranium minerals. However, whilst Olympic Dam ores contain anomalous levels of REE, the concentrations are not sufficiently high to justify economic extraction at present (2022).
The estimated indicative REE composition of the Olympic Dam deposit are as follows (Ehrig, K., AusIMM Geoscience Society Webinar: Olympic Dam, the world's largest IOCG deposit - But what about rare earth elements?, 28 July, 2023):
Rare Earth Elements in ppm = La - 1000; Ce - 1515; Pr - 129.7; Nd - 343.9; Sm - 34.11; Eu - 10.62; Gd - 20.29; Tb - 2.55; Dy - 13.76; Ho - 2.68; Er - 7.74; Tm - 1.13; Yb - 7.47; Lu - 1.03; Y - 73.51. LREE - 3053; HREE - 110; TREE - 3163.
Rare Earth Oxides + Y in ppm = La - 1173; Ce - 1774; Pr - 151.8; Nd - 401.1; Sm - 39.55; Eu - 12.30; Gd - 23.38; Tb - 2.94; Dy - 15.79; Ho - 3.07; Er - 8.85; Tm - 1.30; Yb - 8.50; Lu - 1.18; Y - 93.35. LREO+Y - 3575; HREO+Y - 135; TREO+Y - 3710.
Within a total mineral resource of 11.68 Gt of ore (BHP Annual Report, 2022), this equates to 43.333 Mt of contained TREO+Y. However, Ehrig (2023) points out, these grades are in excess of one order of magnitude less than grades reported by Wang et al (2015) for mines in production, with the exception of alluvial and ion-absorption mineralisation.
The REE minerals in the olympic Dam deposit are typically <9 µm in size, and after grinding to 80% passing -75 µm, the REE minerals remain significantly locked within gangue minerals.

Mineralised alkaline ultramafic to mafic dykes - as detailed above in the Geology section, alkaline ultramafic to mafic dykes, lavas and possibly sills comprise ~30 vol.% of the Olympic Dam Breccia Complex, including those interpreted as the protoliths of heavily hematite altered rocks (Ehrig, 2022, from unpublished mass balance calculations). Johnson and McCulloch, (1995) estimated the ultramafic to mafic rocks accounted for >13 vol.% of the breccia complex, but contained ~50% of the copper in that complex. By their nature, the protoliths of these mafic/ultramafic rocks are highly chemically reactive and typically contain chalcocite-bornite dominant sulphide assemblage, irrespective of where they intrude the broader zonation pattern. Mineralisation may be hosted within the mafic protoliths, generally accompanied by brecciation and strong hematite alteration (Figs. 2 and 3) and, or be concentrated along the margins of dykes that are otherwise barren. Some of the highest copper and gold grades (i.e., >4 wt.% Cu and >2 g/t Au) across the deposit are hosted in intensely hematite altered mafic/ultramafic dyke protoliths (Clark and Ehrig, 2019).
Shallow epithermal mineralisation has been encountered nearer surface and distal to the Olympic Dam Breccia Complex. Examples illustrated in Corriveau et al. (2022) include hematite-fluorite-sulphide 'pisoliths' supported by a hydrothermal siderite cement, interpreted to represent sub-surface hydrothermal boiling textures whereby vugs are infilled from the outside, inward. Another example is of chaotic low-temperature 'epithermal' quartz veins found at a level interpreted to represent a near-surface setting. These are interpreted to represent precipitation resulting in the pressure-temperature reduction and consequent boiling of fluid that escaped from the early overpressured hydrothermal system responsible for the Olympic Dam Breccia Complex.
Two ore types have been delineated within the Olympic Dam Breccia Complex (Ehrig and Clark, 2020), namely a:
• bulk tonnage, low grade, >0.6% Cu equivalent, disseminated chalcopyrite(-pyrite) body of ~10 Gt of ore, represented by the Open Pit Resource as listed in the BHP 30 June 2022 published resources report (see below). This mineralisation is reasonably homogenous and spread throughout the complex, with a slight increase in grade towards the centre of the deposit; and a
• lower tonnage, high grade, >2% Cu equivalent, occurring as a series of structurally focussed, discrete ore shoots that have a strongly anisotropic and variable geometry, as well as a restricted, but similarly anisotropic halo of ~1% Cu disseminated sulphides. This mineralisation is hosted in the fluorite-rich hematite breccias containing bornite-chalcocite, with lesser chalcopyrite. The ore shoots are controlled by: i). zones of intense brecciation accompanied by a high rate of fluid-rock interaction; ii). chemical dissolution > physical brecciation, and iii). local dilatant zones developed at structural intersections.
Satellite Mineralisation
A distal body of satellite mineralisation has been outlined ~4 km NE of the main Olympic Dam deposit. It has dimensions of ~2 x 0.5 km in plan, and 800 m in vertical extent, and lies on a broadly coincident gravity and magnetic geophysical anomaly down-strike from the regional scale Jubilee Fault. The alteration mineralogy and geochemical signature are similar to that seen at the margins of the Olympic Dam Breccia Complex. Alteration comprises hematite-muscovite imposed on the Roxby Downs Granite, visible as red-stained K feldspar that is associated with sparse copper-(iron)-sulphide mineralisation that rarely exceeds 1 wt.% Cu (Kontonikas-Charos et al., 2017). Relict magnetite-apatite-quartz alteration is overprinted by hematite-sericite alteration, although a broad magnetic susceptibility anomaly is evident throughout the satellite mineralised zone, likely reflecting deeper, peripheral alteration linked to the broader Olympic Dam mineral system (Corriveau et al., 2022).

GEOPHYSICAL EXPRESSION

Geophysical data, including deep seismic, magnetotelluric (MT), gravity and magnetic, reflect the 3D setting of the Olympic Dam deposit from a crustal to deposit scale.
Continental-scale seismic tomography datasets indicate the boundary between the lithospheric mantle and underlying asthenosphere is thickest at 180 to 220 km to the west of Olympic Dam, beneath the central Gawler Craton (Kennett et al., 2013), where the depth of the Moho is ~54 km. The latter shallows towards the north, related to offsets across north-dipping trans-crustal faults. To the east, immediately to the SW of the Olympic Dam deposit, the Moho is at depth of ~39 km and becomes slightly deeper to the NE where it is imaged at ~42 km (Drummond et al., 2006). Seismic reflection data shows an offset of the Moho beneath Olympic Dam with slightly thicker crust to the north of the deposit, interpreted to reflect a large crustal-scale fault zone.
Two- and three-dimensional magnetotelluric (MT) modelling to a depth of 60 km has been undertaken on data obtained from traverses across and along the eastern margin of the Gawler Craton, including the Olympic IOCG Province. This modelling revealed a number of domains, as follows (after Heinson et al., 2018): Hematite breccia with sulphides

• A conductive surface layer, corresponding to Neoproterozoic cover sequence of the Stuart Shelf, overlain by transported Tertiary sediments. This layer is similarly evident in reflection seismic sections. The cover thickness around the Olympic Dam deposit is shown to be ~300 m, which is consistent with drilling evidence, but thickens significantly into a deep-rifted basin to the NE, with sedimentary thickness of >2 km.
• Archean to Palaeoproterozoic Gawler Craton crust to the SW is characterised by high resistivity, >1000 Ωm values in MT data, and by high seismic reflectivity, to a depth of >60 km.
• The Palaeo- to Mesoproterozoic crust to the NE has similar high resistivity characteristics to that of the Archaean detailed above, and dips to the NE.
• Between the two high resistivity (low conductivity) domains of the previous two points, there is a 50 km wide zone of low-resistivity, <10 Ωm, that is imaged along the margin of the Archaean Gawler Craton at a depth extending from 15 to 40 km in the mid to lower crust. This highly conductive lower crustal zone is clearly evident in both the long-period MT responses and vertical field anomalies. This conductor is closely coincident with a region of low seismic reflectance, and is centred directly beneath the deep-rifted Neoproterozoic basin to the north east of Olympic Dam. This suggests a zone of crustal rheological weakness. The low seismic reflectance defines a zone where the pattern of reflectors seen to the SW, NE and at shallower levels, fades into a bland area lacking well defined texture. The least resistive, ~1 Ωm part of this domain (in MT data), has an upper termination at ~15 km depth which corresponds to the brittle-ductile transition. Further interpretation of the data, including stations to the east and west of the major NNE-SSW traverse, showed that this major conductivity anomaly was strongly developed at a depth of 25 km, but did not persist to 50 km below the surface. This 3D model showed Olympic dam on the SW gradient of the conductivity high, which is centred to the NE. However, an image that included deeper data in Reid (2019) shows this major conductivity high reappearing to depths of ~100 to 150 km, below the Moho, where it is offset to the east. The shallower and deep conductivity highs are linked by a saddle of marginally lower conductivity. The Moho also appears to be offset by ~3 km, NE-side down, across this composite conductivity high domain, suggestive of a large transcrustal scale fault zone (Reid, 2019). This fault would appear to confirm the similar interpretation of seismic tomography mentioned above. In plan, the resistivity model at an upper lithospheric mantle depth of 55 km, derived from AusLAMP magnetotelluric data (Thiel et al., 2016), shows this deep conductivity high to underlie the Olympic IOCG Province where it is strongly developed to the NE of Olympic Dam, extending to below Prominent Hill and Cairn Hill in the NW. To the south, it decreases in intensity, but then redevelops over a narrower width to continue below Carrapateena (Reid, 2019).
• Three narrow, enhanced conductivity/low-resistivity, ~100 Ωm 'fingers' branch from the top of the major conductive domain at the brittle-ductile transition depth of ~15 km, and extend to the base of the sedimentary cover layer detailed in the first point above. Heinson et al. (2018) state these low-resistivity branches "align remarkably with the spatial location of the major IOCG-U mineral deposits"/occurrences at Wirrda Well, Olympic Dam and Vulcan that are adjacent to the MT transect. The most intense is that below Olympic Dam. These same features are also apparent in processed seismic section as distinct breaks in reflectance horizons and destruction of seismic texture, with an overall low reflectivity. The same authors note however, that these low-resistivity features are not spatially aligned with significant mapped crustal faults.
This pattern of resistivity lows and bland seismic data described in the last two points appears to overprint the Archaean basement and Palaeo- to Mesoproterozoic rocks of the crust up to and including the Hiltaba suite, but are overprinted by the response from the Meso- to Neoproterozoic cover sequence. They are therefore assumed to be related to the early Mesoproterozoic magmatic/hydrothermal event that affected the eastern Gawler Craton, related to channeling of magmas and/or associated metasomatic fluids into the middle and upper crust. Wise and Thiel (2020) suggest this conclusion is supported by the observation that the lower crustal conductors can be traced to surface expressions of magmatic suites such as the A-type Hiltaba granites and Gawler Range Volcanics in the Olympic Domain. It has been suggested that the high Fe content as magnetite in these A-type granites may in part explain elevated conductivities (Yang and Emerson, 1997). However, anomalous fluorine enrichment is regionally associated with these granites and accompanying alteration minerals, and is also a characteristic of the Olympic Dam deposit (McPhie et al., 2011; Xing et al., 2019). Xing et al. (2019) discussed the potential role of fluorine as a mobilising agent, in the form of the highly corrosive hydrofluoric acid, to dissolve and transport U and REE, as well as increasing silica solubility, thus promoting porosity and enhancing fluid pathway connectivity on both a crustal and deposit scale. Skirrow et al. (2018) have suggested the deep, below-Moho zone of high conductivity and bland seismic character, may reflect the presence of a peridotite mantle that has undergone earlier subduction related metasomatism, and been altered to hydrated olivine (e.g., clinohumite), garnet and phlogopite (which may bear fluorine). Experimental studies by Li et al. (2017) have endorsed fluorine as a mechanism for producing high conductivities. Wise and Thiel (2020) conclude that F/Fe-rich magmatic/hydrothermal fluid flow and consequent alteration, would therefore also explain the apparent destruction of seismically reflective lower crust and any associated hydrothermal alteration could leave behind minor mineral phases with enhanced conductivity, e.g. sulphides and F-rich silicates such as phlogopite.
It is therefore possible that the deep, highly conductive domains and shallower, more restricted branches in the Olympic IOCG Province may reflect pathway of fluids and magmas responsible for the large scale transfer of iron oxides, Cu, Au and fluorine from reworking and partial melting of a fertilised lithospheric mantle to the upper crust.

On a deposit to district scale, the geophysical expression of the Olympic Dam deposit comprises a 17 mGal gravity anomaly that is closely coincident with an ~1600 nT aeromagnetic anomaly (Rutter and Esdale, 1985; Esdale et al., 2003) that also closely coincides with the outline of the resource. The bulk dry density of the Roxby Downs Granite and lithofacies within the Olympic Dam Breccia Complex has been shown to steadily increases from 2.60 g/cm³ in the outer 'unaltered' granite, inward towards the centre of the deposit, where it is >4.50 g/cm³. The magnetic susceptibility is directly correlated with the magmatic and hydrothermal magnetite content of the host rock and mineralisation, which varies significantly across the deposit. The major rock units have magnetic susceptibilities as follows: the relatively unaltered Roxby Downs Granite ~25 to 50 SI units; altered and weakly brecciated Roxby Downs Granite <1 SI units; Gairdner dykes 100 to 300 SI units; typical hematite-rich breccias ~0.5 to 55 SI units); magnetite-rich breccias >100 SI units; and barren hematite-rich breccias <1 x 10^-3 SI units.

EVOLUTION and DISCUSSION

This section discusses an interpretation of the evolution of the Olympic Dam deposit, predominantly as an initial multistage event straddling 1590 Ma, possibly partially reactivated 100 m.y. later, with further additions and modifications distributed over the following more than 1 billion years, as it was influenced by tectonic events and related fluid circulation regimes.

Summary
The following is a thumbnail sketch of the proposed evolution of the Olympic Dam iron oxide copper-gold-silver-uranium deposit that is discussed in more detail below, including references to sources. The deposit, which was formed at ~1590 Ma in the early Mesoproterozoic, is hosted by the ~1593 to 1592 Ma Hiltaba Suite Roxby Downs Granite and co-magmatic Gawler Range Volcanics, both of which were cooled and solidified prior to mineralisation. The Olympic Dam deposit is interpreted to be the result of the interaction of a number of fluids that originated at different levels, from the lithospheric mantle to the shallow crust. The principal fluid was initiated by the partial remelting of lower lithospheric mantle that had previously been fertilised by subduction during the ~1850 Ma Cornian and 1730 to 1690 Ma Kimban orogenies which had also produced the extensive Donington Suite granitoids. This partial melting is interpreted to be due to a mantle plume, or delamination and detachment of the lower lithospheric mantle. It produced mafic to ultramafic magmas and fluids, the latter in pressure solution with the former, which rose to be ponded as a large magma chamber at a level of neutral buoyancy in the lower crust. Fractional crystallisation of the mafic/ultramafic magmas was the source of the mafic facies of the Hiltaba Suite/Gawler Range Volcanics; while the oxidised felsic/silicic magmas were produced by partial melting of lower crust by the ascending mafic-ultramafic magmas and the underlying plume, and in part by mixing with fractional crystallised products of the mafic magmas. The apex of the major mafic/ultramafic magma chamber was at a depth of ~15 km, near the brittle-ductile transition. Segregation within the chamber led to the accumulation of fluids and magmas as a reservoir in its apex. Both the ponded fluids and magmas were saturated with volatiles, iron, Cu, U, Au, Ag and REE, but limited sulphur. This reservoir was tapped by the major transcrustal ENE-WSW Jubilee Fault, which focussed both metal and volatile, saturated fluid (Fluid A) and mafic to ultramafic magmas along an intersection with a prominent NNW-SSE structure. The reactive metal and fluid saturated mafic to ultramafic magmas, which are estimated to occupy ~30% of the Olympic Dam Breccia Complex, released and precipitated Cu sulphides internally and, when reacting with fluids, on their margins. Fluid A, which was confined and overpressured, inflated and expanded faults and related stress-field initiated fractures at the structural intersection to facilitate ingress of the strongly corrosive fluorine rich fluid to produce a replacive breccia. This breccia, the Roxby Downs Breccia Complex, was developed between 1 and 2 km below the surface, and relatively passively stoped up- and outward in the freshly cooled, brittle, Roxby Downs Granite and Gawler Range Volcanics. In doing so, the hot Fluid A cooled sufficiently to form early albite-actinolite ±magnetite alteration, overprinted by a biotite-magnetite-K feldspar IOA style assemblage at temperaures of >400 to 600°C; then magnetite-biotite with low grade 0.1 to 0.3% Cu pyrite-chalcopyrite at ~450 to 350°C. This temperature range was sufficiently low to increase the Cu-saturation of Fluid A and release enough Cu (and other metals) to combine with the limited sulphur in the fluid, but retained significant excess Cu in solution in the Fe²⁺ oxidised, hot fluid. This was stage 1 mineralisation, which was probably stimulated by periodic relief of overpressure afforded by movement on the associated fault system. However, the breccia complex then breached the Gawler Range Volcanics into the 'bedded clastic facies'. This resulted in fluid connectivity to be established between Fluid A and the very much cooler, oxidised and sulphate rich Fluid B resident in the continental red-bed OD Basin. This led to the loss of overpressure and produced a hybrid Fluid A-B that had two affects: i). addition of sulphate; and ii). cooling of the combined fluid from the hot Fe²⁺ magnetite-stable conditions of Fluid A, to a cooler, heavier, Fe³⁺ hematite stable, but still acid and volatile rich Fluid A-B. While the reduction in temperature reduced the solubility of Cu, this was compensated by the increase in the degree oxidation, and hence the metal remained in solution. The cooler, heavier Fluid A-B began to sink in the centre of the breccia complex through brine reflux, and established a 'donut-shaped' convection cell, with the hot Fluid A rising on the margins of the complex and the cold Fluid A-B sinking in the centre. The oxidised, acid, Cu-bearing Fluid A-B leached Cu from, and oxidised the stage 1 mineralised upper breccia complex as it descended. However, the interface with the ascending, less oxidised, hot, Cu-bearing Fluid A produced a redox and thermal front. At this front, Fluid A was cooled and Fluid A-B was reduced. They became Cu-saturated to release the metal in both, and precipitate it by overprinting stage 1 sulphides, as well as combining sulphate in the latter with Cu from both fluids to form new sulphides. This was stage 2 mineralisation which took place at 250 to 150°C. The spent Fluid A-B was then re-heated and pushed aside to join Fluid A, rising on the complex margin to close the convection cell. As this process proceeded, the 'hole in the donut' was progressively oxidised and leached to become the barren, hematite saturated HEMQ breccia. Fluid flow through the HEMQ breccia was optimum up to a composition of ~50% hematite, but then waned and was finally terminated when this had increased to 75% as its permeability was occluded, with only fracture controlled quartz and barite veins subsequently forming in the outer margins of that breccia. The result was a barren hematite breccia underlain and fringed by high grade Cu-rich and S-poor sulphides, zoned downward and outward from the redox front from chalcocite → bornite → chalcopyrite → pyrite, as explained in more detail below. During the stage 2 mineralisation, a proportion of fluids A and A-B escaped into the 'bedded clastic facies' and through faults to the surface to accommodate continuing inflow of Fluid A from below. Late in the system development, the inflow of the latter waned, the system cooled to Fe³⁺ hematite stable conditions and a hematite regime overprinted most of the magnetite altered column to depth. The occlusion of HEMQ and waning of inflow of Fluid A represents stage 3. The emplacement of the Olympic Dam deposit was entirely within an extensional regime, which in the lower to middle crust accommodated the emplacement of the voluminous deep magma chamber, and at surface produced the horst and graben setting of the OD red-bed Basin, into which the 'banded clastic facies' was deposited. This extension was responsible for the expansion of the OD Basin during development of the Olympic Dam Breccia Complex and mineralisation. It also led to downfaulting against a major fault-set resulting in the rotation of section of the 'banded clastic facies' and its underlying basement into the breccia complex. Alteration and mineralisation of these rocks had commenced prior to tilting and continued afterwards. As presaged above, following the cessation of stage 3 alteration, the deposit was periodically subjected to deformation, intrusive activity and fluid flow events that resulted in its partial erosion and preservation, and the addition and remobilisation of mineralisation, extending for more than a billion years into the Lower Cambrian. The latter, in particular, appears to have resulted in the redeposition, and deposition of significant additional, uranium below the Neoproterozoic unconformity.

The host to the Olympic Dam deposit, the Olympic Dam Breccia Complex, was developed within the undeformed, A-type, Roxby Downs Granite, which was emplaced at ~1593.87 ±0.21 Ma as part of the regionally extensive Hiltaba Suite magmatism (Cherry et al., 2018). The Roxby Downs Granite is interpreted to have been intruded into the overlying, broadly comagmatic, but marginally older, Gawler Range Volcanics. The latter were deposited at ~1594.73 ±0.30 Ma (Cherry et al., 2018) over the Palaeoproterozoic Wallaroo Group and ~1850 Ma Donington Suite granitoids. Both the A-type Roxby Downs Granite and bimodal, but predominantly felsic, Gawler Range Volcanics, which are anomalously rich in fluorine (McPhie et al., 2011), were emplaced in an intracontinental, extensional regime and represent a ~1595 to 1545 Ma large igneous province (Jagodzinski et al., 2021).
Mantle to crustal scale systematics - The hosts, alteration and mineralisation of the Olympic Dam deposit are part of a coherent and interrelated system of early Mesoproterozoic magmatism, hydrothermal alteration and deformation that extends for an up to 500 km width, from the western Gawler Craton to eastern Curnamona Province (Budd and Skirrow 2007; Skirrow et al., 2018; Reid 2019). According to Wade et al. (2019), such an extensive system requires the input of a vast amount of energy, which, on such a scale, can only be delivered into the crust from the mantle (e.g., Walter 2014). A study aimed at gaining an insight into the character of the mantle below the Gawler Craton is reported in detail in two papers by Wade et al. in 2019 (referenced below). This study was based on mafic rocks collected from the vicinity of significant IOCG and gold deposits/prospects, coming mainly from the mafic suite of the Gawler Range Volcanics, and filtered to exclude samples that have undergone crustal assimilation. These data indicate a mantle source for the mafic suite of the Hiltaba/Gawler Range large igneous province that is dominantly sub-crustal lithospheric mantle (SCLM) with varying magma compositions. These variations are interpreted to represent a vertically and laterally heterogeneous SCLM. In general, these rocks are characterised by enrichment in HFSE (high field strength elements) and LREE (light rare earth elements), negative Nb-Ta-Ti anomalies, Th/Nb ratios of <0.25 and variable Nd isotopic compositions. The geochemical data from samples taken from the Olympic IOCG Province form a distinct group, characterised by moderate to high Cr and Ni; general enrichment in HFSE including Zr, Nb, Th and Y; and moderate LREE enrichment. Sm-Nd isotopic compositions from the IOCG system samples are generally the most evolved. These characteristics are consistent with an SCLM source region that has been modified/metasomatised by subduction processes and was therefore hydrated, or partly hydrated mantle (e.g. Goodenough et al., 2002; Sandeman et al., 2003; Cai et al., 2013; Pearce et al., 2015; Condie and Shearer, 2017). Subduction and metasomatism have been recognised as being important processes that generate metal enrichment in the mantle (Richards 2009; Groves et al., 2010; Tassara et al., 2017; Xu et al., 2017), resulting in residual siderophile and chalcophile-rich zones within the lower SCLM, a potential viable source of Cu and Au.
Wade et al. (2019) indicate the isotopic composition of these magmas from the eastern Gawler Craton suggest subduction may have been as early as the late Neoarchaean, or crust of that age reworked during the Palaeoproterozoic. However, the current craton margin parallels the inferred active margin of the ~1850 Ma Cornian Orogeny and the suture between the two cratonic masses involved that were amalgamated into a single block during the 1730 to 1690 Ma Kimban Orogeny. The remnants of these two cratonic masses now constitute the Gawler Craton and Curnamona Province respectively, separated by the Neoproterozoic Adelaide Rift Complex which reactivated the Palaeoproterozoic suture during the Neoproterozoic. These two earlier collisional subduction events may well have left a legacy fertilised sub-crustal lithospheric mantle below the Olympic IOCG Province.
With minor variations on that proposed by Wade et al. (2019), a model is suggested, involving:
i). a rising mantle plume generated the extensional regime that accompanied the Hiltaba/Gawler Range large igneous province event; or alternatively, regional extension led to delamination and detachment of the lower asthenospheric mantle and the decompression melting and upwelling of asthenospheric mantle;
ii). the hot plume/upwelling asthenosphere is envisaged to have preferentially partially melted the more susceptible hydrated, metasomatised lithospheric mantle (SCLM), but also to a lesser degree the less- or un-metasomatised, anhydrous adjacent SCLM;
iii). this partial melting produced hydrous and oxidised mafic/ultramafic magmas, pressure mixed with voluminous 'hydrous fluids' as a magmatic volatile phase, containing H2O, Cu, Au, Fe and F stored in SCLM that had been metasomatised during the earlier Cornian and Kimban orogenies;
iv). oxidised felsic magmas of the Hiltaba Suite were produced by partial melting of lower crust by the mafic-ultramafic magmas and plume, and in part by mixing with fractional crystallised phases of the mafic magmas;
v). the mafic to ultramafic magmas ascended into the crust to their level of neutral buoyancy, where they differentiated and cooled, accompanied by boiling that released the pressure-mixed magmatic volatile phase which collected in the roof of the magma chamber, at or near the ductile-brittle transition.
vi). the magmatic volatile phase-fluid, composed of hydrous, oxidised, Fe- and F-rich fluids containing elevated Cu-Au, was tapped where the roof of the magma chamber was breached by major transcrustal brittle structures (e.g., the Jubilee Fault) and ascended to contribute to the formation of IOCG mineralisation at shallow levels.
NOTE: The higher SG mafic to ultramafic melts were mostly trapped at their level of buoyant equilibrium in the lower crust, while lighter crustal contaminated and volatile rich mafic magma (e.g., lamprophyres), and the lower SG felsic magmas continued to shallower levels. In contrast, the overpressured fluids followed available, mainly structural dilational pathways towards the surface followed by light, fluid- and gas-saturated mafic to ultramafic magmas.
Mantle to deposit scale connection - The vertical and lateral heterogeneity of the SCLM below the Olympic IOCG Province is also supported in geophysical data by the presence of domains with high conductivity (seen in MT data), and coincident 'bland seismic character' within more extensive background zones of low conductivity and well developed seismic texture. These geophysical anomalies illustrate a convincing connection between the lower lithospheric mantle and the Olympic Dam deposit, and are described in the main Geophysical Expression section above. They are also consistent with the model proposed in the previous paragraph (after Wade et al., 2019). Regionally extensive, deep, conductive and bland seismic domains, which persist over lengths of hundreds of kilometres, are interpreted to reflect zones of Palaeoproterozoic (or possibly late Neoarchaean) subduction imposed metasomatism within the lower lithospheric mantle, that have been reworked and partially melted during the early Mesoproterozoic by a major heat source, as detailed above. The re-melting of the earlier subduction imposed metasomatic zones is interpreted to have produced extensive, hydrated, Cu-Au-F fertile altered zones whose conductivity is influenced by water (as OH^-), fluorine and iron minerals. Upward, within the upper lithospheric mantle and lower crust, similar conductive and bland seismic domains are interpreted to reflect alteration generated by the passage of these fertile mafic to ultramafic magmas and large volumes of un-mixed magmatic volatile phase fluids. Such magmas and fluids are envisaged to have inherited metals and volatiles stored in the fertile, pre-existing subduction metasomatised lithospheric mantle. It is also likely that the highly corrosive, fluorine-rich, metalliferous magmatic volatile phase fluids thus generated, scavenge further metals, particularly Fe oxides, from the lithospheric mantle and crust during their ascent, as well as creating this alteration. The upper conductive and bland seismic domain is consistent with ponding of the mafic to ultramafic magmas and magmatic volatile phase fluids, up to the ductile-brittle transition in the crust at ~15 km, as described above. The pregnant fluids were then focussed into brittle, structurally controlled conduits, emanating from that ponded reservoir. These conduits are interpreted to be reflected by more restricted, but well-defined, 5 to 10 km wide 'fingers' of the same conductive-bland seismic character emanating from the considerably more extensive underlying domain. These fingers are clearly shown to pass up into the Olympic Dam and other IOCG deposits and significant prospects within the Olympic IOCG Province as described above in the Geophysical Expression section. The 'finger' that underlies Olympic Dam is one of the most intensely conductive and has a width at the Neoproterozoic unconformity of the same order as the Olympic Dam Breccia Complex into which it passes.
The ~50 km² and 500 to >1000 m thick Olympic Dam Breccia Complex embraces and hosts the Olympic Dam deposit. The texture, contact relationships, distribution and non-stratified character of the breccias of the complex have been interpreted to be consistent with subsurface fragmentation of already solid granite, caused by a combination of tectonic and hydrothermal processes (Oreskes and Einaudi, 1990; Reeve et al., 1990; McPhie et al., 2011; Ehrig et al., 2021). The deposit and the complex appear to be centred on the intersection of the ENE to NE-SW trending regional Jubilee (or Masher's) Fault and the postulated pre-mineral NNW-SSE fault that is inferred to also control the elongation of the mineralised zone. The core of the complex is dominated by breccias containing clasts of the Roxby Downs Granite the Gawler Range Volcanics, and the younger 'bedded clastic facies'. The latter had been deposited at ~1590.97 ±0.58 Ma (Cherry et al., 2018) in a fault bounded sedimentary basin, overlying both the Roxby Downs Granite and Gawler Range Volcanics at Olympic Dam (McPhie et al., 2016). This basin is most likely one of a number of interconnected early fault controlled depocentres in the evolution of the ~1575 to 1400 Ma Cariewerloo rift basin into which the Pandurra Formation (dated at one location as ~1424 Ma Beyer et al., 2018) was deposited. A wedge of sandstone interpreted to belong to the Pandurra Formation is structurally interleaved with the 'bedded clastic facies' within the Roxby Downs Breccia Complex. The provenance of granitoid detritus in the 'bedded clastic facies' is interpreted to most likely be the Roxby Downs Granite. This would require cooling and partial exhumation of the granite prior to ~1591 Ma (Cherry et al., 2018), most likely during the ∼3 m.y. gap between granite emplacement and deposition of the 'bedded clastic facies'. The southeastern lobe of the deposit is also characterised by the presence of substantial wedge shaped blocks of 'bedded clastic facies' rocks that have been tilted to near vertical dips and down-faulted into the Olympic Dam Breccia Complex in an extensional stress field. These blocks have vertical and horizontal dimensions of the order of 1 to 2 km and hundreds of metres in thickness. They have been brecciated, altered and mineralised and are overprinted by the most intense HEMQ hematite-quartz breccia alteration phase, at least in part following down-faulting. However, McPhie et al. (2016) also consider it likely that the 'bedded clastic facies' were deposited while the Olympic Dam hydrothermal system was active. This implies long-lived hydrothermal activity most likely involving repeated episodes.
There is evidence within the Olympic Dam Breccia Complex that the interpreted deep mantle derived fluids i). are oxidised and rich in Fe, as indicated by the oxidation and intense iron oxide alteration and replacement of the hosts; and ii). contain anomalous fluorine and are very corrosive with a low pH. The latter is consistent with the high relative fluorine content of altered rocks and strong chlorite and sericite to advanced argillic alteration of the breccia complex. It is also emphasised by the vast amounts of silica produced by the alteration of silicate minerals that has been removed during alteration, as well as the corrosion and dissolution of free silica/quartz from the host rocks (as outlined in the Alteration section above).
The development of the Olympic Dam Breccia Complex is envisaged as follows, largely after McPhie et al. (2011) and Cherry et al. (2017):
• Deposition of the bimodal Gawler Range Volcanics at ~1594.73 ±0.30 Ma over the Palaeoproterozoic Wallaroo Group and ~1850 Ma Donington Suite granitoids, was followed by intrusion of the Roxby Downs Granite at ~1593.87 ±0.21 Ma, all within an extensional regime;
• Soon after deposition and cooling to a brittle state, a regional deep seated fault system propagated into the Roxby Downs Granite, and initiated structurally controlled brecciated at the intersection of major, steep, ENE to NE-SW and NNW-SSE fault sets. These were trans-crustal structures, which tapped the reservoir of cogenetic un-mixed deep crustal fluids and volatile-rich mafic to ultramafic melts ponded near the ductile-brittle transition at a depth of ~15 km. Both the melt and the hydrated, oxidised, iron, fluorine, Cu and Au rich fluids were channelled upwards via the dilational, structurally brecciated fault intersection. These overpressured fluids, ascended through the fractures of the mechanical breccia, replacing wallrocks, creating an upward migrating replacive breccia, the nascent Olympic Dam Breccia Complex. It is suggested, that while some of the fluid pressure may have been periodically relieved to surface by movement on the major faults, the fluid system was still predominantly under overpressure. This overpressure facilitated hydraulic brecciation and inflation (natural fracking) along weaknesses imposed by the fault system stress field, allowing the maximum access of fluids to develop the replacive breccia.
• Displacement on these fault systems in an extensional stress field also initiated changes in the topography at surface to produce local uplift and erosion from horsts and deposition in an associated graben (the 'OD Basin'), in a 'basin and range setting. As a consequence, the Gawler Range Volcanics and then the Roxby Downs Granite were exposed and by ~1590.97 ±0.58 Ma supplied detritus into the OD Basin to deposit the red and green mudstones and sandstones of the 'bedded clastic facies', whilst distal eruptions contributed tuffaceous mudstone. During this period, the breccia complex had propagated into the base of the Gawler Range Volcanics, as had the volatile and gas-rich ultramafic to mafic dykes that had accompanied the deep hydrothermal fluids in their ascent. At this stage the hydrothermal brecciation dominated over the structural fracturing.
• Continued extension initiated the Woodall and Woodall Jr. faults, and deepening of the OD Basin graben. The pronounced downthrow on the northern side of these faults led to the flat lying lower sections of the semi-consolidated 'bedded clastic facies' strata being tilted and collapsing 'from either side', first as a down-warp, then as fault controlled wedges, into the accommodation space generated by the extension. Meanwhile, at the basin floor level, debris flow conglomerates were deposited from the steeper fault controlled scarps of the graben as part of the ongoing deposition. During this stage, the breccia complex encroached into the lower 'bedded clastic facies' and the hot mantle derived hydrothermal fluids mixed with cold basinal brines. It is likely that this breakthrough from the confining granite and volcanic rocks, into the relatively shallow, permeable, cooler and more oxidised red bed sedimentary basin may have resulted in a change from the earlier higher temperature, pressure and less oxidised conditions.
• As the overall regime described above evolved, fluid flow from below continued to produce progressively more intense alteration and 'replacive brecciation' affecting the Roxby Downs Granite, Gawler Range Volcanics and 'bedded clastic facies', including brecciation, dissolution of contained volatiles and internal alteration and mineralisation of the mantle sourced mafic to ultramafic dykes that accompanied that fluid upward from the apex of the parental magma chamber at depth.
• As the overlying OD Basin matured in a continuing extensional setting, it was amalgamated with other localised interconnected grabens, and become part of the developing, broader, ~1575 to 1400 Ma Cariewerloo rift basin, in which, possibly after a hiatus, the more widespread Pandurra Formation was deposited. Late tectonic activity along the Woodall faults resulting in slivers of that sandstone also collapsing into the breccia complex with underlying units of the upper 'bedded clastic facies'. The down-faulting of the deposit and deposition of the 400 to 2000 m thick Pandurra Formation in the ~420 x 170 km Cariewerloo red bed basin, concealed and preserved the deposit from erosion. Dating of uraninite within the deposit has yielded a small number of dates at ~1.44 Ga, which are taken to reflect minor remobilisation related to the Late Kararan Orogeny that involved reactivation of major regional shear zones and commenced the inversion of the Cariewerloo Basin (Ehrig and Clark, 2020).
The evolution of the breccia complex as outlined above, the inter-relationship of lithologies, structural framework, and the character of the rocks, all suggest the development of the Olympic Dam Breccia Complex involved a relatively long lived, progressive, pulsed, but reasonably passive process, at shallow depths, most likely punctuated by localised small scale phreatic bursts. This is implied by:
i). Depth - although the hydrothermal system has roots suggested to extend to the brittle-ductile transition at ~15 km depth, and from there to the mantle, the formation of the deposit is indicated to have principally taken place at between 500 and 2000 m below the contempory surface. The lower limit, taken here is the lowest >5 wt.% Fe granite to hematite breccia lithofacies which occurs at a depth of <1.5 km below the base of the >500 m thick 'bedded clastic facies', the top of which was the sedimentary basin floor. Interpreted exhalative chemical precipitates within the 'bedded clastic facies' and sub-surface boiling textures in the deposit are additional evidence the system evolved at a depth of <1 to 2 km and interacted with the overlying active OD Basin (Corriveau et al., 2022). Subsequent structural attenuation may have thinned this interval which is underlain and fringed by a greater thickness of less intense alteration on the margins of the system.
ii). Temporal extent - the development of the deposit is inferred to have continued over much of the interval in which the 'bedded clastic facies' were deposited, and continued as they collapsed into the breccia complex, as outlined previously.
iii). Passive setting - during this period, apart from along scarps on the basin margin where conglomerate-breccia wedges formed, deposition of the directly overlying 'bedded clastic facies' produced a thick sequence of laminated to finely bedded mudstone and sandstone, albeit gently to moderately slumped, even while the breccia complex was encroaching into the lower members of that facies <500 m below. A dynamic, venting breccia zone would be expected to produce a chaotic mélange.
iv). Replacive brecciation - in the early stage of brecciation, clasts of the breccia are observed to comprise horses of the protoliths bounded by fault induced fracture margins which were then progressively replaced inwards from those fractured margins until only their ghosts remain (Ehrig and Clark, 2019)
v). Collapse - clasts within the breccias appear to largely be of in situ origin, composed of lithologies either from the same level at which the breccia is formed, or are 'exotic', predominantly from stratigraphically higher units. This implies continued replacive brecciation, followed by structural and solution collapse brecciation. The latter could be explained by volume reduction through alteration and removal in solution of components such as silica by the corrosive hydrothermal fluids and hydrolytic alteration. This volume reduction, coupled with accommodation within the extensional regime which also resulted in physical collapse of large wedges of 'bedded clastic facies' from higher in the system. The extreme alteration in the core of the system produced disaggregation of the protolith and degradation of clasts to progressively finer particle sizes. This is envisaged as culminating in fluidisation of the matrix, allowing mixing of clasts during a long lived continuation of repeated fluid flow pulses. However, clasts of early magnetite alteration are seen, overprinted by chalcopyrite ±pyrite mineralisation, and preserved within a low temperature K-Fe altered granite breccia (e.g., Corriveau et al., 2022). These large clasts are interpreted to have been transported vertically upwards and incorporated into parts of the shallower alteration facies, although they may represent relicts of early in situ prograde magnetite alteration, overprinted by later retrograde hematite-sulphide facies.

Deposition of Mineralisation
The transport and deposition of copper in almost all deposit types is governed by both the metal's availability and solubility. There are four main variables that singly or in combination determine the solubility of copper, resulting in saturation and partial to total release from solution: i). temperature and pressure, where the metal remains in solution until one or both are sufficiently reduced, e.g., in porphyry Cu deposits; ii). pH, whereby Cu is held in solution at low pH until the transporting fluid is neutralised/diluted, e.g., in supergene blankets; iii). Eh - where Cu remains in solution in an oxidised regime, until more reduced conditions are encountered, e.g., sediment hosted Cu deposits; iv). interaction with reactive rocks/minerals/fluids/gases that extract Cu from solution; these include in order of declining reactivity: sulphates/sulphides/and 'other sulphur anions'; carbonates; mafic minerals; feldspars. It can be argued that all of these four controls were active in parts of the Olympic Dam mineralised system.
Among all of the detail described in this summary, a number of key observations seem important when considering the potential processes involved in the formation of the deposit, as detailed, with source references, above. They are:
• The sulphur content as sulphides, which decreases upwards from deep in the Roxby Downs Breccia Complex to the base of the hematite-quartz-barite HEMQ breccias, or the Neoproterozoic unconformity peripheral to the flanks of the latter. This corresponds to an upward transition from pyrite (~53% S) → chalcopyrite (~35.0% S) → bornite (~24.5% S) → 'chalcocite' (~20% S) from the base to top respectively. However, as Cu grades in the chalcocite zones are 2 to 3 times those that are predominantly chalcopyrite, the total sulphur content is not significantly lower. Never-the-less, overall, the sulphur content of the system increases downward and persist with depth, implying it was largely derived from below.
• The copper content is generally highest adjacent to and within ~200 m below the lower and lateral contact of the HEMQ breccias, and gradually decreases away from that contact, corresponding to the mineralogical transition from chalcocite → bornite → chalcopyrite → chalcopyrite-pyrite → pyrite, i.e., from minerals that contain the most to the least copper per atom of sulphur. This implies that, at least in part, Cu was supplied from above.
This conclusion is based on experience in other mineralised systems, particularly hypogene sediment hosted deposits and supergene enrichment (e.g., of porphyry Cu deposits). These systems display a transition across a redox boundary, from an oxidised setting to reduced conditions with a zonation (in part or in full) from chalcocite → bornite → chalcopyrite → pyrite. This zonation can, in those systems, be shown to represent the passage of Cu-bearing fluids from the oxide setting towards the outer pyrite zone, with the highest grades and Cu:S ratios adjacent to the redox boundary and the lowest distal to that interface. This is related to the properties of Cu and should equally apply to IOCG systems. It reflects the progressive addition of Cu to the sulphide lattice, which is most pronounced from closest to the source, and least, distal to that same source. This process is usually accompanied by the expulsion and sometimes associated deposition of iron oxides.
• Two ore types have been recognised within the Olympic Dam Breccia Complex. These are:
i). High grade, >2% Cu equivalent, lower tonnage (~1 Gt) ore, occurring as a series of structurally focussed, discrete ore shoots that have a strongly anisotropic and variable geometry, as well as a restricted, but similarly anisotropic halo of ~1% Cu disseminated sulphides. This mineralisation is hosted in fluorite-rich hematite breccias containing bornite-chalcocite, with lesser chalcopyrite. It is accompanied by intense brecciation and a high rate of fluid-rock interaction where chemical dissolution is > physical brecciation; and
ii). Low grade, >0.6% Cu equivalent high tonnage (~10 Gt), disseminated chalcopyrite (-pyrite) ore. This mineralisation is reasonably homogenous and spread throughout the complex, but principally below the high grade ores, with a slight increase in grade towards the centre of the deposit.
• The hematite saturated, sulphide free, 'barren', hematite-quartz-barite (HEMQ) breccias are the most intensely altered and oxidised rocks in the Olympic Dam Breccia Complex, albiet formed at the lowest temperatures. They represent a regime in which copper is the most soluble and hence, if present may not be precipitated. These breccias, however, are generally closely flanked laterally and downward by the highest grade, highest Cu:S ratio minerals (chalcocite-bornite), which decrease in both grade and Cu:S ratio away from a broad zone adjacent to that contact (Reynolds, 2000). Uranium mimics the distribution of copper, whilst significant LREE mineralisation, and some of the higher grade gold is concentrated between the uraniferous bornite-chalcocite mineralisation and the ‘barren’ HEMQ core of the deposit, straddling either side of the contact with the latter.
• Two different geochemical associations, or 'IOCG signatures', have been recognised within the IOCG mineralisation of the Olympic Dam Breccia Complex, with separate, distinct, spatial distributions: i). Cu-U3O8-Se-S, the main economic and closely associated elements, have a concentric zonation, outward and downward from the base of the HEMQ breccia. ii). Au-W-Mo-Sb-As, which forms a vertical, >1800 m deep elongate column, still open at depth, in the southeastern lobe of the deposit, spatially coinciding with hematite (>5% Fe) and volcanic breccias within the Olympic Dam Breccia Complex. It is especially evident at depths below 1200 m, as well as in generally shallow quartz-barite-hematite breccias (HEMQ) and hematite-quartz 'bedded clastic facies' sediments (KHEMQ). This vertical mineralised column is considered likely to be related to the major subvertical WSW-ENE striking fault system acting as a hydrothermal-fluid conduit. However, the concentration and zoning of the first group, while being controlled by the same mineralisation/alteration corridor, appears to overprint and be spatially independent of the Au-W-Mo-As-Sb association, i.e., two separate mineral associations taking advantage of the same conduit.
Although the available interpretations and literature suggest the HEMQ represents the outflow of a spent mineralising hydrothermal fluid following ore deposition (e.g., Ehrig and Clark, 2020), another variation might be considered as follows.

It is suggested the formation of the Olympic Dam IOCG-U-Ag deposit involved the interaction of three fluids, based on observations recorded at Olympic Dam, as described and cited above, and from similar deposits/prospects within the same district:
Fluid A - is the ascending hydrothermal fluid suggested to have been derived from parental fluid generated during partial melting of pre-existing hydrous, subduction metasomatised and fertilised, mantle lithosphere. This parental fluid then ascended, pressure-mixed with mafic to ultramafic magmas, to be ponded near the ductile-brittle transition in the crust, as detailed in the Geophysical Expression section above. The apex of that very large ponded reservoir at ~15 km depth is seen as having been tapped via deep brittle structures that have channelled a segregated fraction of that fluid upwards, accompanied by volatile-laden mafic to ultramafic magmas. As such it is derived from a magmatic mass of hundreds of km³, rather than having unmixed/degassed from magmatic bodies of maybe several km³ within of shallowly below the ore deposit. Such a large system is essential to providing the tonnage of Cu at Olympic Dam. However, because of their common origin, the volatile-rich ~1590 Ma mafic to ultramafic magmas will also carry similar elements and fluids which will unmix/degas at shallower depths and also contribute to the deposit. Fluid A is seen to have been very hot (having deposited a biotite-magnetite-K feldspar assemblage deep in the Olympic Dam Breccia Complex at >400 to 600°C). It was also overpressured, as suggested by the hydraulic opening of a denser array of fractures related to the prevailing stress field, with contemporaneous, inward progressive alteration of the blocks of granite between those fractures. Based on the associated alteration, the fluid had an intermediate to reduced oxidation state, carrying appreciable Fe²⁺, (e.g. as aqueous FeCl2; Crerar and Barnes, 1976; Seyfried and Ding, 1993) which will tend to limit its sulphur carrying content (H2S >> ΣSO4^2-) to relatively low levels (Skirrow, 2022). It is hydrated, rich in volatiles, particularly fluorine, and carried anomalous quantities of Cu, U, Au and Ag, all derived from re-melting of previously subduction fertilised lithospheric mantle. High temperature fluids of this type found within the Olympic Dam district carry 2 to 4 wt.% Cu in some fluid inclusions, which including verified chalcopyrite. If chalcopyrite-bearing inclusions are excluded, these fluids still carry 300 to 600 ppm Cu (Bastrakov, Skirrow and Davidson, 2007).
The sourcing of ore forming fluid from a very large, deep, differentiated parental magma chamber, is directly analogous to what is now understood to produce porphyry Cu deposits, but with differently sourced felsic magmas. In porphyry systems, the segregated fluids, sourced from the apex of the deep parental magma chamber, are accompanied by fluid saturated porphyries derived from the same part of the batholith which contribute mineralisation in the deposit via the pre-solidus A-veins, while the bulk of the ore is provided by the deeply sourced fluid, occurring as B, C, D, etc., veins cutting the solidified porphyry and country rock. As such, the mineralised ~1590 Ma mafic/ultramafic rocks at Olympic Dam would be analogous to the intrusive stocks of a porphyry system.
The main point of difference between porphyry and IOCG deposits, is that in the former, other than the early A-veins, the bulk of mineralisation occurs as discrete veins, whilst even the disseminated sulphides are in now closed microfractures. This is the product of overpressured fluids facturing a brittle host. In contrast, IOCG deposits often maintain a porosity within the matrix that accommodates the passage and deposition of fluids, with veins at Olympic Dam largely confined to the fractured granites and volcanic rocks external to or as horses within the breccia complex, and late following the occlusion of porosity and solidification (e.g., the carbonate veins).
Fluid B - is a cool, oxidised basinal brine, accumulated in an arid to semi-arid intra-continental red-bed basin; the OD Basin into which the 'bedded clastic facies' were deposited. This basin is interpreted to be fault controlled, related to the same structural network that controlled the ingress of Fluid A. The 'bedded clastic facies' are overlain by the Pandurra Formation deposited in the succeeding Cariewerloo rift basin. It is likely that the 'bedded clastic facies' were deposited in one of an interconnected network of 'basin and range' type grabens and horsts of the type that usually precede the development of rift basins. This is supported by the recognition of other similar sedimentary sequences of the same age distributed along the eastern margin of the current Gawler Craton (Ehrig and Clark, 2020). Bastrakov, Skirrow and Davidson (2007) describe fluid inclusions that suggested oxidised, lower temperature, SO4-rich brines formed in these basins and were involved in the development of the hematite alteration in the deposits of the Olympic Dam IOCG Province. This is also supported by the observation by Oreskes and Einaudi (1992) that the hematite associated ore-bearing breccias were deposited by fluids of significantly lower temperature and lower δ¹⁸O values than Fluid A.
Fluid A-B - a hybrid fluid resulting from the mixing of fluids A and A-B, initially in the apex of the Olympic Dam Breccia Complex, when it finally breached the base of the OD Basin, and then in the permeable sands of the 'bedded clastic facies'. The composition and temperature of this hybrid fluid will vary considerably, depending on its location relative to the two end members.

Based on observations and interpretations in a range of sources, including Skirrow (2022), Reid (2019), Dmitrijeva et al. (2019), Schlegel et al. (2017), Ehrig et al. (2017), Krneta et al. (2017), Bastrakov, Skirrow and Davidson (2007), Haynes et al. (1995), Oreskes and Einaudi (1992), and other references listed previously, the Olympic Dam deposit may have initially formed as follows:
• Fluid A was tapped from the apex of the deep mafic to ultramafic magma reservoir and channelled up the regional, transcrustal Jubilee Fault, focussed along its intersection with a NNW-SSE fault that was to control the elongation of the Olympic Dam deposit. This occurred following solidification of the Roxby Downs Granite, after ~1593 Ma.
• Brecciation and alteration - the fluid column within the fault controlled channel hydraulically transmitted the deep llithostatic pressure to shallower levels to open and inflate fractures initiated by the prevailing stress field (natural 'fracking'). This allowed ingress of the highly corrosive hydrofluoric acid rich Fluid A into those fractures to decompose and alter the inter-fracture Roxby Downs Granite. This initiated the development of the Olympic Dam Breccia Complex as both a mechanical and replacement breccia. In so doing, it appears to have produced an early albite-actinolite ±magnetite alteration, overprinted by a biotite-magnetite-K feldspar assemblage at temperaures of >400 to 600°C. During the same period, the 'bedded clastic facies' were being deposited within a basin controlled by the fault regime related to the ingress of fluids. As such those faults may have periodically relieved the overpressure and allowed escape of Fluid A to shallower levels, to percolate within, sub-surface, and into at surface, the sediments of the OD Basin, particularly the hematitic mudstone and sandstone of the KHEMQ unit, implanting an IOCG signature.
• Upward migration and stage 1 mineralisation - the Olympic Dam Breccia Complex is interpreted to have generally passively expanded upward and outward through the Roxby Downs Granite and into the overlying Gawler Range Volcanics. During this period, alteration continued to evolve, producing potassic alteration, progressing to assemblages of magnetite-fluorapatite-ankerite-calcite and siderite-chlorite-quartz. This alteration was accompanied by weak Cu sulphides and uranium, possibly to grades of 0.1 to 0.3% Cu, typical of the magnetite rich early stage mineralisation found elsewhere in the Olympic IOCG Province. The precipitation of sulphides during this stage was initiated by progressive cooling, and took place in the magnetite stability field of ~450 to 350°C as the fluid ascended, likely enhanced by periodic release of overpressure via the major faults. However, at these temperatures, Cu was only mildly exceeded saturation, insufficient to promote significant exsolution, and in conjunction with the low levels of available sulphur in the fluid, left substantial excess Cu in the residual Fluid A. This stage of mineralisation was restricted to a pyrite → pyrite-chalcopyrite assemblage and is interpreted to correspond to the Au-W-Mo-Sb-As 'IOCG signature' seen from the lowest to highest development of alteration within the Olympic Dam Breccia Complex. This signature is also seen in the red sandstone facies of the 'bedded clastic facies' where fluids may have escaped to surface or near surface in the faults bounding the OD Basin, as discussed above.
• Breakthrough into the OD basin and transition to hematite stability - the Olympic Dam Breccia Complex is interpreted to have breached the capping Gawler Range Volcanics into the 'bedded clastic facies'. This led to the mixing of fluids A and B, a reduction in pressure regime, lowering of temperature and possibly an increase in the upflow of Fluid A, as well as the addition of solutes. This mixing had two main consequences: i). a sharp decrease in temperature caused the acidic, Cu-bearing Fluid A to transit from high temperature Fe²⁺ magnetite-stable conditions to Fe³⁺ hematite stability, enhanced by mixing with the cold and strongly oxidised Fluid B; ii). the addition of solutes from Fluid B, particularly SO₄^2- and brines generated in the red bed basin. The mixing of these two fluids formed the hybrid Fluid A-B. While the reduction in temperature decreased the solubility of Cu, this was compensated by the increase in oxidation state, maintaining the metal in solution.
• Convection - the hybrid Fluid A-B was a much cooler brine, and as such, was denser than Fluid A. As a consequence, it began to reflux, sinking in the centre of the breccia complex where it displaced ('pushed aside') the hotter ascending Fluid A, to form a NNW-SSE elongated, 'donut shaped' convection cell. The sinking Fluid A-B caused the rising hot Fluid A below to be displaced laterally on the underside of the donut, towards the outer margins of the Olympic Dam Breccia Complex, thus widening the shallow complex on its margins. It then ascended around the outer and upper side of the donut to encounter Fluid B above, formed the cool Fluid A-B, and then migrate toward the centre of the system, where it refluxed, down through the 'hole in the donut'. The cooler, sinking, Fluid A-B in the centre of the 'donut' was eventually heated at depth by the ascending Fluid-A, and began to rise again on the margins of the complex with the latter.
• Stage 2 mineralisation - the Cu-rich and SO₄^2--bearing Fluid A-B was strongly oxidised and incapable of releasing the residual Cu inherited from Fluid A. As it sank through the 'hole in the convection donut', it was sufficiently acidic to supplement its load by stripping additional metal from breccias of the upper stage 1 mineralised Olympic Dam Breccia Complex, whilst oxidising the leached host rock, converting magnetite to hematite. This process initiated the formation of the HEMQ hematite-quartz-barite breccia which became increasingly hematite-rich, and grew to fill the 'hole in the donut'. However, at the interface with the upwelling, hot, less oxidised, ferrous iron (Fe²⁺) stable, magmatic-hydrothermal Fluid A, the strongly Fe³⁺ oxidised Fluid A-B was reduced. Under these conditions it could not retain its contained metal load in solution, which was then released. Multiple processes, or combinations thereof, took place at this interface, which represented a downward and outward advancing redox front. Because this redox front was between two fluids, its location varied due to temporal difference in the relative flow rate of the two fluids and the porosity of the country rocks at any one point. As a consequence it was irregular. The processes and reactions that deposited Cu sulphides at this front included:
i). Released Cu from Fluid A-B reacted with pre-existing stage 1 pyrite and chalcopyrite to progressively produce chalcopyrite, bornite and eventually chalcocite. However, stage 1 sulphides continued to be deposited by the ascending Fluid A, below and lateral to the stage 2 mineralisation, beyond the immediate influence of the descending Fluid A-B.
ii). Exsolved Cu and SO₄^2-, both from Fluid A-B, combined to precipitate Cu sulphides;
iii). Remaining Cu in the ascending Fluid A, upon being cooled by Fluid A-B, reacted with the SO₄^2- of the latter, as well as stage 1 mineralisation to form further sulphides;
iv). Reactive mafic to ultramafic (lamprophyre) dykes, which are estimated to occupy ~30% of the Olympic Dam Breccia Complex, caused Cu to be released from both fluids A and A-B, in the presence of a sulphur source, to produce Cu sulphides, and from Fluid A alone, deeper in the mineralising system. These magmas, which were likely also saturated with metal and volatiles will have a). contributed to the heat budget; and b). released and precipitated additional Cu sulphides internally and, when reacting with fluids A and A-B, on their margins.
Gold is also carried in fluids A and A-B, and has a solubility that is similarly strongly influenced by Eh, but to a lesser degree than Cu. As a consequence, it is deposited inboard of Cu, straddling the oxidation front.
These stage 2 reactions took place at temperatures of 250 to 150°C (Skirrow, 2022, Skirrow, 2010, Bastrakov, Skirrow and Davidson 2007). They were most prevalent within a transition zone of 100 to 500 m outboard of the advancing redox front, with the Cu grade rapidly increasing ahead of that front, to a peak characterised by a chalcocite-bornite assemblage. Past this zone of maximum Cu deposition, both the amount of available, unreacted Cu and SO₄^2- contributed by Fluid A-B declined. As a result both the Cu grade and Cu:S ratio also progressively declined, with the transition from chalcocite → bornite → chalcopyrite-pyrite → pyrite. From the Cu grades, Cu:S ratios of mineralisation, and the general upward decline in S content, it would appear Fluid A-B only made a relatively minor sulphur contribution, with the bulk being from the stage 1 sulphides that were partly overprinted and rimmed by later phases, and partially dissolved and redeposited. However, Bastrakov, Skirrow and Davidson (2007) note that sulphur in both stage 1 and stage 2 alteration assemblages from deposits and sub-economic prospects in the Olympic Dam district was derived either from cooling magmas and/or crystalline igneous rocks carried by relatively oxidised fluids (ΣSO4^2- ≈ ΣH2S, δ³⁴Ssulphides from –5 to +2‰) or from crustal sedimentary rocks (δ³⁴Ssulphides from +5 to +10‰).
Ehrig and Clark (2020) note that the high grade (stage 2) mineralisation is often aligned parallel to the late flat (listric), extensional fault set in the upper parts of the deposit (Fig. 3). This may reflect earlier, now annealed, syn-mineral, fault set related to the extensional regime operating during mineralisation.
Stage 2 mineralisation was accompanied by lower temperature retrograde hydrolytic hematite-sericite-chlorite-carbonate alteration, and strong silica depletion. Fluid A-B continued to descend, altering the country rock, before being diluted and heated by Fluid A, then rising on the margins of the complex as part of the 'donut shaped' convection cell. During stage 2, ingress of Fluid A continued to contribute metal and other solutes to the system. This had to be balanced by upward or outward fluid loss responsible for the shallow sub-surface boiling textures and epithermal veins in sections of the system. The stage 2 mineralisation was responsible for the Cu-U-Au-Ag 'IOCG signature' that accompanies and overprints the Stage 1 Au-W-Mo-Sb-As signature.
The redox front represented by the lower and lateral margins of the leached, sulphide-free HEMQ breccia was not static, but advanced outward and downward with the continuing flow of Fluid A-B, scavenging high grade stage 2 chalcocite-bornite mineralisation in its path, leaving barren HEMQ in its wake, and reinforced mineralisation ahead; basically sweeping a progressively higher grade wave of mineralisation ahead of its advance. Silica leached during the hydrolytic alteration and transported by the spent Fluid A-B within the convection cell was deposited in veins on the margin of the Olympic Dam Breccia Complex and within the HEMQ breccia forming a silica-hematite cap, whilst SO₄^2- contributed by Fluid B was involved with the deposition of sulphates, particularly barite in the HEMQ breccia core.
High grade stage 2 mineralisation gradually decreased in intensity as the deposition of hematite progressively increased within the HEMQ breccia occluding its porosity. Fluid flow through the HEMQ breccia was optimum up to a composition of ~50% hematite, but then waned and was finally terminated when this had increased to 75%, with only fracture controlled quartz and barite veins subsequently in the outer margins of that breccia.
• Stage 3 alteration/mineralisation - which took place in the waning stages of the mineralised system. The temperature of Fluid A was progressively reduced, and the hematite to magnetite stability front descended, whilst the stage 2 convection cell closed down. This resulted in hematite further overprinting earlier alteration assemblages and the cessation of significant mineralisation.

'Bedded Clastic Facies'
Prior to deposition of the stage 2 mineralisation, the Olympic Dam Breccia Complex had broken through the roof of the Roxby Downs Granite and the Gawler Range Volcanics to encroach up into the 'bedded clastic facies'. As detailed above, fluid connectivity was then established between the breccia complex and the permeable units of the 'bedded clastic facies', allowing mixing between fluids A and B and limited deposition of mineralisation within the sedimentary rocks of the OD Basin. The rocks of the 'bedded clastic facies' were mostly above or lateral to the main mineralising domain, and as such have not been brecciated and altered to the degree of the bulk of the rocks of the Olympic Dam Breccia Complex. Prior to, and following this event, it is likely that tectonic activity allowed periodic escape of Fluid A from the breccia complex up major structures, such as the Masher's and Woodall faults, resulting in ingress into permeable units in the OD Basin above the breccia complex, and hot springs on the basin margin feeding fluid and solutes into the basin. In both cases this would have resulted in hematitic alteration, particularly of the KHEMQ hematite-quartz-rich sandstone and red mudstone, and imparting the IOCG signature on that unit. It is uncertain as to why the green sandstones and mudstones of the KASH unit were not similarly altered and mineralised, although its different provenance may have been related to a change in the structural setting that changed the topography at surface and also the access to escaping fluids. During the formation of the Olympic Dam deposit, the continued extensional regime led to the tilting and collapse of large slabs of the 'bedded clastic facies' into the Olympic Breccia Complex, while the hydrothermal system remained active and to continued alteration and mineralisation, as on the margin of the KASH unit and within the deep KHEMQ fault slabs. Fluid activity appears to have continued sporadically, or as a reactivated event, after or during deposition of the 1575 to 1490 Ma Pandurra Formation.

Post-mineral Evolution
The Olympic Dam Breccia Complex and the Olympic Dam deposit were buried below the 400 to 2000 m thick Pandurra Formation of the ∼42 500 km² Cariewerloo red bed basin. According to Giles, Betts and Lister, (2004), the Cariewerloo basin extension culminated in the separation of the North and South Australian cratons across a WNW-ESE line to the NE at ~1.5 Ga. These were followed by a series of tectonic and hydrothermal events. that led to the structural and mineralogical reorganisation of the deposit, as follows:
• The inversion Cariewerloo Basin was followed by a long period of erosion and sedimentation. However, a scattering of dates evident in the uraninite of the deposit that are poorly constrained between 1300 and 1100 Ma, straddle the major regional ~1200 to 1160 Musgravian deformational event. These are suggested to be related to the sinistral fault dismemberment/displacement of the deposit that is evident on the main geological plan above. On a global scale, this activity coincided with the end of the break-up of the Columbia Super-continent and the initiation of assembly of Rodinia (Ehrig and Clark, 2020).
• There is some structural activity within the Olympic IOCG Province during the assembly of Rodinia between 1.1 and 0.9 Ga, including the possible uplit and exposure of the Pandurra Formation at ~1.1 Ga. The next significant event marked the onset of the break-up of that super-continent, namely the major ~825 Ma Gairdner Dyke Swarm large igneous province. Members of this swarm intrude the deposit with minor hydrothermal modifications and alteration related to their margins.
• During the Cryogenian period of the Neoproterozoic, the 750 to 700 Ma Sturtian and 660 to 636 Ma Marinoan glaciations levelled the topography above the Olympic Dam deposit and formed the Neoproterozoic unconformity surface on the Stuart Shelf. At Olympic Dam, this erosive event removed much of the Pandurra Formation, 'bedded clastic facies' and Gawler Range Volcanics that had protected and preserved the deposit for a billion years, and exposed the mineralisation. In the immediate deposit area, these eroded units were only preserved as down faulted remnants within the Olympic Dam Breccia Complex. However, this glacial levelling and exposure of the deposit was accompanied by deposition of a thick glacial and then shelf sequence, which continued to protect the deposit from further erosion to the present. The unconformity is only underlain by a thin palaeoregolith.
• As detailed above in the Mineralisation section, a significant period of U mineral formation occurred over the interval from 0.7 to 0.5 Ga, broadly consistent with the occurrence of non-euhedral uraninite that is part of a uranium assemblage that differs from those dated at ~1950 Ma. This second, younger assemblage is interpreted to imply widespread concomitant modification of precursor mineralisation, and possibly growth of new, uranium and sulphide minerals. This may be related to the unconformably overlying shelf sediment sequence that directly overlies the deposit, and/or the onset of the ~515 to 485 Ma Delamerian Orogeny during the assembly of the Gondwana block of the transient Pannotia supercontinent. Activity at the unconformity is indicated by alteration and sulphide mineralisation described above in paragraph vii). of the Carbonate Minerals section of the ALTERATION division.
• To the east, this Neoproterozoic to Phanerozoic cover sequence passes into the thicker, north-south oriented major 870 to 500 Ma Adelaide Rift Complex. This rift complex opened along the Palaeoproterozoic suture and separated the Mawson Craton from the Curnamona Province to the east. Later, during the Cretaceous, the separation of Antarctica resulted in the Australian segment of the Mawson Craton becoming the present Gawler Craton. In this sense, the location of the Olympic Dam deposit and the Olympic IOCG Province has progressed from near a cratonic margin during the Palaeoproterozoic, to intra-continental in the Late Palaeoproterozoic to the Early Neoproterozoic, to craton margin again from the Mid-Neoproterozoic.

RESERVES and RESOURCES

At the end of 1989, after commencing mining operations in mid-1988, reported resources and reserves (Reeve et al., 1990) amounted to:
Measured + indicated resource = 450 Mt @ 2.5% Cu, 0.6 g/t Au, 6.0 g/t Ag, 0.8 kg/tonne U3O8,
Inferred resource = 2000 Mt @ 1.6% Cu, 0.6 g/t Au, 3.5 g/t Ag, 0.6 kg/tonne U3O8,
Proved reserve = 13 Mt @ 3.0% Cu, 0.3 g/t Au, 10.2 g/t Ag, 1.1 kg/tonne U3O8,
Proved gold reserve = 2.3 Mt @ 1.6% Cu, 3.6 g/t Au, 2.9 g/t Ag, 0.3 kg/tonne U3O8.

At December 2004, published (BHP Billiton, 2005) reserves and resources were:
Proved+probable reserves totalled 761 Mt @ 1.5% Cu, 0.5 g/t Au, 0.5 kg/tonne U3O8,
within a total resource of 3810 Mt @ 1.1% Cu, 0.5 g/t Au, 0.4 kg/tonne U3O8.

At 30 June 2012, the published resources (BHP Billiton, September, 2012) amounted to:
Measured resource = 1474 Mt @ 1.03% Cu, 0.35 g/t Au, 1.95 g/t Ag, 0.30 kg/tonne U3O8,
Indicated resource = 4843 Mt @ 0.84% Cu, 0.34 g/t Au, 1.50 g/t Ag, 0.27 kg/tonne U3O8,
Inferred resource = 3259 Mt @ 0.70% Cu, 0.25 g/t Au, 0.98 g/t Ag, 0.23 kg/tonne U3O8,
Total resource = 9576 Mt @ 0.82% Cu, 0.31 g/t Au, 1.39 g/t Ag, 0.26 kg/tonne U3O8.
This resource includes a total proved + probable reserve of:
629 Mt @ 1.76% Cu, 0.73 g/t Au, 3.36 g/t Ag, 0.57 kg/tonne U3O8.
At the same date, the separate non-sulphide gold resource was 364 Mt @ 0.75 g/t Au, comprising:
Measured resource = 73 Mt @ 0.85 g/t Au; Indicated resource = 255 Mt @ 0.73 g/t Au; Inferred resource = 36 Mt @ 0.70 g/t Au.

At 30 June 2015, the published resources (BHP Billiton Annual Report, 2015) amounted to:
Measured resource = 1.330 Gt @ 0.96% Cu, 0.40 g/t Au, 2.0 g/t Ag, 0.29 kg/tonne U3O8,
Indicated resource = 4.610 Gt @ 0.79% Cu, 0.32 g/t Au, 1.0 g/t Ag, 0.24 kg/tonne U3O8,
Inferred resource = 4.120 Gt @ 0.71% Cu, 0.24 g/t Au, 1.0 g/t Ag, 0.25 kg/tonne U3O8,
Total resource = 10.060 Gt @ 0.78% Cu, 0.30 g/t Au, 1.0 g/t Ag, 0.25 kg/tonne U3O8.
This resource includes a total proved + probable reserve of:
484 Mt @ 1.95% Cu, 0.74 g/t Au, 4.0 g/t Ag, 0.59 kg/tonne U3O8.
Stockpile - 44 Mt @ 0.99% Cu, 0.51 g/t Au, 2.0 g/t Ag, 0.37 kg/tonne U3O8.
At 30 June 2015, a separate non-sulphide gold resource was 283 Mt @ 0.84 g/t Au, which was not reported in 2015.

At 30 June 2017, the published resources (BHP Annual Report, 2017) amounted to:
Measured resource = 1.460 Gt @ 0.96% Cu, 0.41 g/t Au, 2.0 g/t Ag, 0.30 kg/tonne U3O8,
Indicated resource = 4.680 Gt @ 0.79% Cu, 0.34 g/t Au, 1.0 g/t Ag, 0.25 kg/tonne U3O8,
Inferred resource = 3.920 Gt @ 0.71% Cu, 0.28 g/t Au, 1.0 g/t Ag, 0.24 kg/tonne U3O8,
Total resource = 10.100 Gt @ 0.78% Cu, 0.33 g/t Au, 1.0 g/t Ag, 0.25 kg/tonne U3O8.
This resource includes a total proved + probable reserve of:
508 Mt @ 1.99% Cu, 0.72 g/t Au, 4.0 g/t Ag, 0.58 kg/tonne U3O8.
Low grade Stockpile - 37 Mt @ 1.13% Cu, 0.51 g/t Au, 3.0 g/t Ag, 0.36 kg/tonne U3O8.

At 30 June 2022, the published resources (BHP Annual Report, 2022) amounted to:
Open Pit
Measured resource = 3.300 Gt @ 0.58% Cu, 0.32 g/t Au, 1.0 g/t Ag, 0.19 kg/tonne U3O8,
Indicated resource = 3.220 Gt @ 0.56% Cu, 0.24 g/t Au, 1.0 g/t Ag, 0.19 kg/tonne U3O8,
Inferred resource = 3.070 Gt @ 0.58% Cu, 0.23 g/t Au, 1.0 g/t Ag, 0.20 kg/tonne U3O8,
Total resource = 9.950 Gt @ 0.57% Cu, 0.26 g/t Au, 1.0 g/t Ag, 0.19 kg/tonne U3O8.
Underground
Measured resource = 0.800 Gt @ 1.55% Cu, 0.61 g/t Au, 3.0 g/t Ag, 0.46 kg/tonne U3O8,
Indicated resource = 0.750 Gt @ 1.42% Cu, 0.53 g/t Au, 3.0 g/t Ag, 0.43 kg/tonne U3O8,
Inferred resource = 0.18 Gt @ 1.40% Cu, 0.61 g/t Au, 3.0 g/t Ag, 0.39 kg/tonne U3O8,
Total resource = 1.730 Gt @ 1.48% Cu, 0.58 g/t Au, 3.0 g/t Ag, 0.44 kg/tonne U3O8.
TOTAL Underground + Open Pit Resource = 11.680 Gt @ 0.70% Cu, 0.31 g/t Au, 1.3 g/t Ag, 0.23 kg/tonne U3O8.
This resource includes a total proved + probable underground reserve of:
585 Mt @ 1.82% Cu, 0.67 g/t Au, 5.0 g/t Ag, 0.56 kg/tonne U3O8, plus
Low grade Stockpile - 42 Mt @ 0.81% Cu, 0.33 g/t Au, 2.0 g/t Ag, 0.27 kg/tonne U3O8.

Annual Production:
2005-06 totalled 204 300 tonnes of Cu, 3.34 t Au, 27.49 t Ag, 3936 tonnes U3O8.
2008-09 totalled 194 100 tonnes of Cu, 3.36 t Au, 29.14 t Ag, 4007 tonnes U3O8.
2011-12 totalled 192 600 tonnes of Cu, 3.66 t Au, 28.21 t Ag, 3885 tonnes U3O8.
2014-15 totalled 124 500 tonnes of Cu, 3.26 t Au, 22.52 t Ag, 3144 tonnes U3O8.
2020-21 totalled 205 300 tonnes of Cu, 4.54 t Au, 22.2 t Ag, 3267 tonnes U3O8.
2021-22 totalled 138 400 tonnes of Cu, 3.72 t Au, 23.1 t Ag, 2375 tonnes U3O8.

The mine is owned and operated by a subsidiary of BHP Ltd.

Return to top

Prominent Hill ...................... Wednesday 23 November, 2005.

The Prominent Hill iron oxide copper-gold (IOCG) deposit is located approximately 150 km north-west of Olympic Dam and 650 km NNW of Adelaide in northern South Australia (#Location: 29° 43' 14"S, 135° 34' 38"E).

Prominent Hill, Carrapateena, Olympic Dam, Moonta-Wallaroo and Hillside, and all of the other significant known IOCG mineralised systems of the Gawler craton, are hosted within Palaeo- to Mesoproterozoic rocks and are distributed along the eastern edge of the currently preserved craton, to define the Olympic IOCG Province.

The Prominent Hill IOCG deposit was discovered through diamond drilling of a gravity anomaly in 2001 (Carter et al., 2003) and was brought into production in 2009.

Cratonic to Domain-scale Setting

See the Gawler Craton and Olympic IOCGU Province record for a summary of the regional, cratonic setting of the Olympic IOCG Province.

Prominent Hill lies on the southern margin of the Mount Woods Domain (MWD), which is located within the Olympic IOCG Province in the northeast of the preserved Archaean to Mesoproterozoic Gawler Craton. The MWD comprises Palaeo- and Mesoproterozoic metamorphic and igneous rocks that have a well defined aeromagnetic signature. It is characterised by high magnetic and gravity signatures, caused by multiple iron-oxide and mafic rock sources, and its boundaries are sharp and structurally controlled. The MWD encloses a major regional ~75 x 50 km magnetic complex with an overall 'U'-shape, elongated in a NW-SE direction and open to the SE. Outcrop is sparse, with flat lying Phanerozoic cover reaching thicknesses of up to 400 m in places, although it is generally <200 m.

The MWD comprises at least two separate sedimentary successions that have been subjected to one or more amphibolite to granulite facies metamorphic events, three periods of deformation, and two (probably three) episodes of magmatism, as well as a pulse of Neoproterozoic mafic dyke intrusion. Nine basement sub-domains have been defined, within the southern half of the domain within a 40 km radius of Prominent Hill.

The core of the southern half of the MWD is occupied by an 'eye-shaped', east-west elongated, 25 x 10 km kernel of magnetically anomalous rocks, reflecting the lopolithic White Hill Igneous Complex, which is characterised by extremely high amplitude magnetic lineaments on its margins, more subtle concentric magnetic zoning in its centre, and a very pronounced and complex gravity signature. The complex (where drilled) comprises pyroxenite, norite and gabbro, with pronounced layering defined by plagioclase and pyroxene-rich layers, with interleaved disseminated to massive magnetite-ilmenite bands. The gabbros and pyroxenites are hydrous, alkaline and enriched in volatile components, and carry up to 8% modal apatite. The complex is divided by a north-south fault into the White Hill and Joe's Dam sub-domains to the east and west respectively, based on differences in the magnetic pattern. The White Hill Complex is not as well developed in the Joe's Dam Sub-domain. Where drilled, the latter comprises a suite of quartzo-feldspathic gneisses with magnetite rich gneiss intercalations, and rare bands of mafic rocks, supporting the interpretation that the White Hill Igneous Complex is not well represented to the west. The magnetite and hematite breccias and replacement bodies of the Joe's Dam sub-domain are hosted by psammitic, pelitic and gneissic metasediments of higher metamorphic grade than those in the marginal Neptune and Blue Duck sub-domains. The voluminous Hiltaba Suite magmatism (Hiltaba granites and mafic to ultramafic White Hill Igneous Complex) in the area are a potential cause of high temperature contact metamorphism of the pelites (Freeman and Tomkinson, 2010).

The lenses-shaped 20 x 5 km Kennedy's Dam Sub-domain forms much of the northern margin of the White Hill and Joe's Dam sub-domains, and is characterised by relatively low magnetic susceptibility enclosing a few moderate to high amplitude, linear magnetic anomalies. Where drilled, it is composed of a sequence of quartz-feldspar-biotite-magnetite gneisses and 'granulites' with very minor amphibolite and calc-silicate units, intruded locally by coarse grained granite and pegmatite.

The Skylark Sub-domain bounds the Kennedy's Dam Sub-domain to the NW, and is characterised by large intrusions with low to moderate magnetic responses, surrounded by narrow aureoles of higher magnetic intensity related to either hornfelsing or magnetite metasomatism. Granite with actinolite±magnetite veins and pyroxenites, norites and diorites dated at 1587±4 Ma (U-Pb; Jagodzinski, 2005) have been encountered in drill holes, intruding a variety of quartz-feldspar-biotite gneisses, all with accessory magnetite. Metasediments away from the intrusions include quartz-rich meta-sandstone with plagioclase, opaque oxide (magnetite and hematite), schistose biotite, fluorite and tourmaline. Samples of cordierite-garnet-bearing pelite and magnetite psammite have maximum depositional ages of ~1750 Ma respectively ( U-Pb zircon; Jagodzinski et al. 2007), and were intruded by the syn-metamorphic Engenina Adamellite (~1691±25 Ma; Finlay 1993; Daly et al. 1998). Geophysical data suggest the domain may be underlain by a continuation of the White Hill Igneous Complex.
Mount Woods Inlier

Tectonic and geophysical setting of iron oxide related mineralisation in the southern Mount Woods Domain (MWD) - a). RTP aeromagnetic image of the southern MWD showing the sub-domains discussed herein. The sub-domains are based on exploration drilling and geophysical data; b). The interpreted sub-domains of the southern MWD and location of prospects discussed. The prospects correspond to magnetic and/or gravity anomalies. Modified after: Freeman and Tomkinson (2010); Schlegel et al. (2018).

The 50 x 5 km Taurus Sub-domain forms the NE margin the Kennedy's Dam Sub-domain, but persists further to the east, delineating the northern margins of the White Hill and then the Blue Duck sub-domains, and the southern margin of the Ware's Peak Sub-domain to the NE. This terrane, characterised by a series of short strike length curvilinear magnetic features by major discontinuities, is interpreted to represent rocks caught up between a major strike-slip shear couple.

The extensive Ware's Peak Sub-domain occupies a large area in the NE of the MWD and is characterised by a few high amplitude, strike continuous, but folded magnetic linears, and by large areas of diffuse magnetic signature suggesting moderately magnetic and eastward deepening younger cover. Drill holes testing prospects encountered a highly variable suite of magnetite, plagioclase and garnet bearing paragneisses (including graphite rich units), as well as feldspathic and quartz-poor to quartz-rich meta-igneous lithologies (including quartz diorite, monzodiorite, syenite, granite and pegmatite). Other lithologies include metasandstone, iron formation, calc-silicate, skarn, dolomite and marble units. Undeformed granite and gabbroic intrusions have also been identified in drilling. Metamorphic grades range from mid-amphibolite to granulite facies.

The curvilinear, lensoid 2 to 6 km wide and >60 km long Blue Duck Sub-domain extends along the southern margin of the Joe's Dam, White Hill, Taurus and Ware's Peak sub-domains. In the west, it forms the southernmost sub-domain of the MWD. It is thickest in the centre, where it is immediately to the north of the Neptune Sub-domain that contains the Prominent Hill deposit near the contact between the two terranes, and thins on either extremity. In the east, its southern margin is marked by the regional ENE-trending Bulgunnia Fault. It is characterised by narrow, medium amplitude magnetic linears which are continuous for several kilometres or more. Drilling at a number of locations (mostly aimed at magnetic targets), has encountered metamorphosed calcsilicate units with associated magnetite, schistose, recrystallised K feldspar-biotite-chlorite-quartz-scapolite clastic metasediments, iron-rich, pelitic carbonate rock, marble, calc-silicate, magnetite-pyroxene-quartz rock and possible meta-evaporites. These rocks have been tentatively correlated with the 1760 to 1740 Ma Wallaroo Group found elsewhere in the Olympic IOCG Province (Freeman and Tomkinson, 2010).

The elongate Neptune Sub-domain forms the southern boundary of the MWD over an interval of ~30 km, where it fringes part of the southern Blue Duck Sub-domain, separated by the major, south vergent, Southern Overthrust. It has an overall 'V' shape, turning on its eastern extremity from an east-west to WSW trend, along the northern side of the regional Bulgunnia Fault. It includes the , which are up to 4.5 km thick (tentatively correlated on lithological grounds with the ~1590 Ma Gawler Range Volcanics) and the Prominent Hill Mine Sequence. Compared to terrains further to the north, it is characterised by relatively low amplitude linear aeromagnetic anomalies and contains a sequence of lower greenschist facies, relatively undeformed, mafic to felsic volcanic rocks (basalt-andesite-dacite-rhyolite), hematite-cemented quartz conglomerate, sandstone, argillite and dolostone. This sequence coarsens southwards from argillaceous and calcareous rocks into coarse grained siliciclastic rocks. The volcanic component in the structural footwall to the Prominent Hill deposit are basaltic to andesitic in composition, commonly porphyritic and amygdaloidal. Several kilometres to the west, felsic volcanic rocks (dacite to rhyolite) become much more voluminous with 'redrock' hematite dusting of alkali feldspar, compared to the predominantly sericite-chlorite-earthy hematite-leucoxene-carbonate alteration within the basalts and andesites of the mine sequence. Fragmental lithologies are intercalated with the coherent footwall volcanic rocks at Prominent Hill, and include agglomerate, felsic tuff or ignimbrite (Belperio et al., 2006) and volcanic clast conglomerate. They are also intercalated with mafic to intermediate volcanic rocks and hematitic, quartz-feldspar conglomerate and interbedded coarse grained sandstone. The east-west-trending Hanging wall Fault at Prominent Hill has been inferred to separate the hematite-stable sedimentary rocks of the copper-gold mineralised host sequence of the Neptune Sub-domain in the south, from a magnetite-stable hanging wall sequence of chloritic pelite and pelitic carbonate rock of the Blue Duck Sub-domain to the north. Proterozoic basement is overlain by 90 to 150 m of flat-lying Permo-Carboniferous sandstone and diamictite and Cretaceous sandstone and black claystone.

The Danae Hill Sub-domain forms the southern margin of the MWD to the southeast of the regional Bulgunnia Fault. It is characterised by a similar magnetic signature to the Neptune Volcanics, with low to moderate amplitude aeromagnetic linears. Drilling has identified a suite of altered, low-grade metamorphosed, sheared and brecciated basalts, with lesser metasediments and acid volcanic rocks, which collectively suggest a bimodal volcanic suite. These volcanics have a markedly different trace element signature similar to that of the Neptune Volcanics and are inferred to be of late Neoarchaean to Palaeoproterozoic in age.

The Christie Domain, is mainly found to the immediate south of the of, but also includes the MWD Neptune/Blue Duck and Danae sub-domains, and is predominantly composed by latest Archaean to earliest Palaeoproterozoic metasedimentary protoliths, metamorphosed in the earliest Palaeoproterozoic at ~2450 Ma during the Sleaford Orogeny to become the Christie Gneiss of the Mulgathing Complex. The MWD and has been overthrust onto the Christies Domain from the NE, with a major north dipping thrust, the Southern Overthrust marking the boundary (Betts et al., 2003).

In summary, the MWD is predominantly composed of ~1760 to ~1740 Ma Palaeoproterozoic meta-sedimentary and meta-volcanic rocks, tentatively correlated with the Wallaro Group seen elsewhere in the Olympic IOCG Province, including banded iron formations (possibly equivalent to the Middleback Range BIFs further to the south near Whyalla). These successions, which may include inliers of older late Neoarchaean to lower Palaeoproterozoic metamorphic rocks, underwent peak metamorphism at ~1736±14 Ma to amphibolite and granulite facies. They are bounded on the southern margin of the domain, across the south-vergent Southern Overthrust, by younger, low- to mid-greenschist facies, ~1590 Gawler Range Volcanics equivalents and intercalated volcanic and sedimentay breccias of the Neptune Sub-domain, hosting the Prominent Hill ore deposit. All of these rocks have been intruded by the ~1691±25 Ma Engenina Adamellite, the extensive Hiltaba Suite 1584±18 Ma Balta Granite and mafic to ultramafic rocks that include the 1562 ±14 Ma (U-Pb zircon; Allen et al., 2016) White Hill Igneous Complex and similar 1587±3 Ma (U-Pb zircon; Jagodzinski, 2005) leucogabbro near Peculiar Knob.

The presence of Hiltaba-aged zircon, interpreted as metamorphic (Holm, OZCHRON), in quartzites and felsic gneisses suggests that metamorphic grade in the MWD at ~1590 to ~1580 Ma was significantly higher than is typical elsewhere in the Gawler Craton during that period. This, in turn, suggests that the MWD was at deeper crustal levels than the adjacent Olympic, Christie and Wilgena Domains.

These differences in crustal level may be explained by Tiddy et al. (2020) who have interpreted the regional structural architecture of the Mount Woods Domain (MWD) to represent a partially preserved metamorphic core complex that developed during the ~1600 to 1580 Ma Hiltaba Event. The lower plate of the complex is preserved in the central to northern parts of the MWD, bounded to the SW, east and NE by the curved Skylark Fault Zone, and to the north and NW by Cairn Shear Zone and the Panorama Fault Zon to the west. The rocks of this plate comprise the Mount Woods Metamorphics which have a maximum detrital zircon depositional age of ~1860 Ma and were polydeformed and metamorphosed to upper amphibolite to granulite facies during the ~1740 to 1690 Ma Kimban Orogeny. These metamorphic rocks cover a WNW-ESE elongated area of ~60 x 30 km. To the south, the upper plate is composed of a younger succession, which includes the Skylark Metasediments that were deposited at ~1750 Ma. Subsequently, sedimentary and volcanic successions of the Gawler Range Volcanics were deposited into half grabens that evolved during brittle normal faulting of the upper plate caused by the updoming that resulted from the rise of the lower plate. This upper plate brittle deformation is underlain by the Skylark Shear Zone which represents the basal detachment fault separating the upper and lower plate of the core complex. The geometry of normal faulting in the upper plate is consistent with NE-SW extension. Both plates were intruded by ~1795 to 1575 Ma Hiltaba Suite granitic and mafic plutons. These intrusions and half graben hosted Gawler Range Volcanics include the White Hill intrusions and sequences of the Joe's Dam, Kennedy and Blue Duck domains to the SE of the Skylark Fault Zone. Extensive modification of the core complex occurred during the ~1570 to 1540 Ma Kararan Orogeny which is characterised by shear zone formation. Exhumation of the complex was associated with activity along syn-Kararan Orogeny faults, as indicated by ⁴⁰Ar/³⁹Ar biotite cooling ages that show that rock packages in the central MWD cooled to below ~300 ±50°C at ~1560 Ma, which was ~20 m.y. before equivalent cooling in the western and eastern Mount Woods Domain. This assumes all three packages were metamorphosed at the same time. The western MWD part of the upper plate had reached peak metamorphic temperatures of ~750°C at ~1590 Ma (Forbes et al., 2011). Exhumation had commenced along the Southern Overthrust at ~1592 to 1582 Ma (Forbes et al., 2012), and was likely further accommodated by Kararan-aged faults. Exhumation continued until ~1540 Ma, likely accommodated by the Southern Overthrust and possibly by continued activity along the Skylark and Panorama Fault Zones (Tiddy et al., 2020). These observation suggest the Hiltaba Suite intrusives and Gawler Range Volcanics were emplaced under a NE-SW directed extensional regime, rather than NW-SE directed transpression.

Regional Alteration and Mineralisation

Two main iron oxide alteration styles are abundant in the southern Mount Woods and northern Christie domains:
• Magnetite-dominant alteration which generally comprises repeated episodes of magnetite ±phlogopite and/or magnetite-K feldspar-pyrite ±chalcopyrite ±pyrrhotite ±apatite ±titanite ±monazite which overprint both metasedimentary and volcanic rocks, and occurred both during and after brecciation. These alteration assemblages are locally accompanied by low grade Cu, mainly due to minor associated chalcopyrite (Schlegel et al., 2015).
Hydrothermal replacement, veins and breccia-fills of (already brecciated) pelitic and calcareous metasediments and gneisses are abundant along the southern and eastern margins of the White Hill Igneous Complex (Freeman and Tomkinson, 2010; Schlegel and Heinrich, 2015). Proximal to the same igneous complex (see images above for location), the metasedimentary rocks are altered to an assemblage that includes magnetite (with minor hematite) ±biotite/phlogopite ±pyrite ±pyrrhotite ±chalcopyrite ±diopside ±actinolite ±talc ±scapolite (Schlegel et al., 2017). A number of large magnetite accumulations with anomalous copper, gold and uranium associated with this alteration are found adjacent to this intrusive complex, and within 10 km to the west in the Joe's Dam sub-domain, often with associated mafic to ultramafic intrusions - see the Manxman - Joes Dam record for a detailed description. These occurrences are interpreted to have developed by contact metamorphism and the release of magmatic-hydrothermal fluid from the White Hill Igneous Complex (Betts et al., 2003; Freeman and Tomkinson, 2010; Schlegel and Heinrich, 2015). On the basis of textural and paragenetic observations (Schlegel and Heinrich, 2015), this regional magnetite-'skarn' alteration and mineralisation may partly pre-date the economic Cu-Au ore formation at Prominent Hill. Further south and distal from the intrusion, weakly magnetite or hematite altered amphibolite facies calcareous and pelitic (meta-) sediments contain abundant pyrite. The fluids involved in this regional alteration may well have resulted from mingling of the mafic to ultramafic, mantle derived igneous complex and felsic gneisses and/or Balta granites of the Joe's Dam sub-domain to initiate exsolution and expulsion of Fe-rich hypersaline (Na-Ca-K) fluids responsible for the regional alteration and mineralisation observed in the southern Mount Woods and northern Christie domains (e.g., Chen et al., 2010; Velasco and Tornos, 2009).
At the Joe's Dam South East prospect, calcareous beds within a sequence of interbedded limestone-pelite and its brecciated equivalents are pervasively replaced by a magnetite ±pyrite ±quartz assemblage (Betts et al., 2003). This alteration took place in stages, separated by brecciation and renewed replacement of the calcareous breccia matrix by magnetite and pyrite. The prospect lies immediately adjacent to the margin of the White Hill Igneous Complex. Magmatic magnetite banding is found in gabbros on the margin of the igneous complex, whilst miarolitic cavities in gabbro, also on the margins of the intrusion, contain K feldspar-phlogopite-magnetite-quartz ±pyrite ±pyrrhotite. Within the sedimentary wall rock, K feldspar-, quartz- and pyrite-bearing pegmatitic veins cut the deformed interbedded limestone-pelite unit which also contain pyrite in calcareous layers that grade into dominantly magnetite-pyrite ±phlogopite-altered limestone interbedded with pelite. The magnetite alteration grades outward into magnetite-poor but pyrite-rich brecciated calcareous beds within the limestone-pelite sequence (Schlegel et al., 2015; 2017).
• Hematite-dominant alteration overprints undeformed albite- and magnetite-K feldspar altered porphyritic to aphyric basaltic andesitic and andesitic extrusives of the 'Neptune Volcanics', interpreted to be equivalents of Gawler Range Volcanics (Carter et al., 2003; Harris et al., 2013) at the Neptune and Triton prospects, 5 to 6 km SE of Prominent Hill. At Triton, these mafic lavas are variously altered to hematite-chlorite-sericite ±quartz ±kaolinite ±fluorite ±apatite ±barite ±leucoxene, surrounding intense hematite-quartz-apatite-barite alteration (Schlegel and Heinrich, 2015). Early crystallised amphibole is replaced by chlorite, and is overgrown by euhedral plagioclase phenocrysts, that in turn, are progressively altered to sericite and chlorite. The transition to the inner hematite-quartz alteration resulted in pseudobrecciation and near total replacement and destruction of mineral textures in the original protolith. Kaolinite is intergrown with sericite and replaces plagioclase crystals or fills volcanic amygdales. Intense hematite-quartz ±apatite ±barite alteration resulted in replacement of feldspars and sericite-kaolinite-filled amygdales, whilst late calcite fills the remaining pores. Local patches of chalcopyrite and minor chalcocite are found in intensely altered autobrecciated intervals between volcanic flows, and in veins containing hematite, euhedral quartz, K feldspar and pyrite. Veins with finely intergrown quartz-hematite-chalcopyrite-pyrite dominate and K feldspar or sericite occur as vein selvages or in the matrix of local breccias. At ~250 m depths, these veins become increasingly brecciated to form a fracture network that hosts Cu mineralised clasts set in a matrix with two generations of chalcopyrite-pyrite that has associated fluorite, barite and dolomite, indicating that mineralisation is related to small, multiple brecciation events. Degassing of volatiles from the lavas at Triton resulted in vesicular flow tops filled with quartz-hematite-magnetite-fluorite-chalcopyrite and late dolomite (Schlegel et al., 2017).
At Neptune, mafic lavas of the 'Neptune Volcanics' are altered to a magnetite ±K feldspar, chlorite ±fluorite ±apatite ±leucoxene ±barite ±carbonate ±pyrite ±chalcopyrite assemblage that are pervasively altered by magnetite-quartz-pyrite-chalcopyrite ±K feldspar ±calcite ±fluorite. Potassic feldspar is commonly reddened by very fine grained hematite. Four main generations of secondary iron oxides are distinguished at Neptune: i). fine granular magnetite, enriched in basaltic protoliths; ii). coarse granular magnetite, mostly associated with late hydrothermal 'mineralisation', some partly converted to hematite; iii). 'fresh', randomly clustered, bladed hematite; and iv). magnetite, partly to completely replacing earlier, randomly clustered, bladed hematite (mushketovite). Broad intervals of strong magnetite alteration contain anomalous to low grade copper mineralisation with minor zones of higher grade restricted to narrow, matrix-supported, magnetite rich hydrothermal breccias.

PROMINENT HILL DEPOSIT

The Prominent Hill deposit was discovered under ~100 m of cover, and is reflected by a discrete gravity anomaly, the target of the discovery drillhole (Carter et al. 2003). This anomaly corresponds to the hydrothermal iron altered (magnetite-deficient) hematite matrix breccias that host the copper-gold-silver-uranium-cerium-lanthanum ore deposit within the Neptune Sub-domain, and to corresponding palaeotopographic highs. The peak of the gravity anomaly coincides with a mass of massive, barren 'steely' hematite-silica flooded volcanics (on the eastern end of the main ore zone), flanked to the west by a ~2 km long mineralised hematite matrix-supported breccia. The overall gravity anomaly has an east pointing "V" form, with a 1 km long northern arm trending WNW-ESE, and the southern 2 km long limb trending WSW-ENE. The southern limb reflects the main mineralised hematite breccia, with the peak of the anomaly near the hinge, and has no magnetic expression (Belperio et al., 2006).

An associated 750 m long magnetic anomaly coincides with the northern limb of the gravity feature. It is centred ~500 m to the north of the orebody, across the steeply north dipping Hanging wall Fault that separates the ore hosting rocks of the Neptune Sub-domain in the south, from the Blue Duck Sub-domain to the north. This magnetic body, which is within the Blue Duck Sub-domain (Hart and Freeman 2003), reflects a package of magnetite-chlorite-tremolite-phlogopite altered metasomatic 'skarn-like' altered calc-silicate/carbonate-rich metamorphic rocks, intercalated with highly altered, intermediate porphyritic intrusives and chlorite matrix tectonic breccias (Belperio et al., 2006). This package largely comprises a body of massive magnetite accompanied by phlogopite-chlorite-pyrite ±chalcopyrite ±actinolite-tremolite ±serpentine ±talc ±scapolite, known as the 'magnetite skarn' (Belperio et al., 2007; Freeman and Tomkinson, 2010; Schlegel and Heinrich, 2015).

The deposit is characterised by abundant, largely pre-mineralisation breccias, large volumes of hematite-white mica-carbonate altered rock, variation of the dominant Cu mineral species (native copper/ chalcocite/bornite/chalcopyrite), and a distinctive geochemical association with marked enrichment of F, Ba, LREE and U (Belperio et al., 2007; Williams et al., 2017).

Geology
The Hanging wall Fault is a relatively brittle, sinistral structure that dips at 70 to 80°N in the deposit area, but flattens to 60 to 60° a little further west. It is represented by a chlorite breccia zone (Freeman and Tomkinson, 2010; Schlegel and Heinrich, 2015). To the immediate north of the main Prominent Hill deposit, it closely follows the major, south vergent Southern Overthrust. The latter is a domain boundary separating a mid- to upper-greenschist to locally amphibolite facies sequence of calcsilicate and phyllite to the north in the Blue Duck Sub-domain, correlated with the Palaeoproterozoic Wallaroo Group, from the less metamorphosed volcanosedimentary succession of the Neptune sub-domain to the south. To the west of the main Prominent Hill deposit, the Southern Overthrust diverges sharply northward whilst the Hanging wall Fault continues with an east-west trend. The Neptune sub-domain sequence hosts the Prominent Hill deposit in the footwall of the Hanging wall Fault/Southern Overthrust, within a 200 to 300 m thick, <1.75 Ga sequence of weakly to unmetamorphosed shallow marine grading to fluvial carbonate and siliciclastic sedimentary rocks, including conglomerates and sedimentary breccias. This sequence is overturned, dipping steeply to the north, subparallel to the Hanging wall Fault. The Main Host Sequence is structurally underlain, but stratigraphically overlain, by a younger, ~1.59 Ma sequence dominated by basalt, basaltic andesite and andesite, correlated with the Gawler Range Volcanics and then by a red bed sequence. All of these rocks are cut by two generations of mafic dykes (Freeman and Tomkinson, 2010; Williams et al., 2017).

The succession in the Prominent Hill deposit area is a follows, from oldest to youngest (after Freeman and Tomkinson, 2010 and Schlegel and Heinrich, 2015):
Hanging wall metamorphic sequence above the Southern Overthrust.
• Mostly unbrecciated metasedimentary rocks that include carbonate- and chloritic pelite/phyllite and K feldspar-biotite-chlorite-quartz-scapolite clastic metasedimentary rocks, limestone and dolostone, and inferred former evaporites (Freeman and Tomkinson, 2010). They are intruded by undeformed reddish granitoids, mainly quartz-diorite, tonalite and a felsic K feldspar-phyric porphyry which has fine-grained hematite dusting. The intrusions also include a ~1585 Ma dacite porphyry (Belperio et al., 2007). This sequence, which is interpreted to belong to the 1765±27 to 1741±1 Ma (GeoScience Australia) Wallaroo Group, is the host to the 'magnetite skarn'.
- Southern Overthrust and in part Hanging wall Fault and unconformity
Main Host Sequence in the footwall of the Hanging wall Fault/Southern Overthrust, sandwiched between the hanging wall metamorphic and footwall volcanic sequences. These rocks have a maximum depositional age of ~1750 Ma (detrital zircon U-Pb; Bull et al., 2013). The host-rock package protoliths are provisionally interpreted by Bull et al. (2015) as consistent with deposition as poorly sorted, subwave base, clastic turbiditic, debris flow and hemipelagic sediments, with intercalated carbonate reefs, deposited in a shallow marginal marine grading to continental and perhaps mildly evaporitic environment in a magmatically and tectonically active basin (Schlegel and Heinrich, 2015). The succession comprises:
• Shallow marine sequence of interbedded and brecciated, argillaceous dolostone, argillite (breccias), calcareous matrix-bearing and coarse-grained siliciclastic, clast-supported breccias, medium- to fine-grained greywacke, chlorite- and muscovite-cemented subarkose, and laminated and massive limestone and dolostone, as well as quartz-feldspar conglomerate and volcanic-clast conglomerate south of the deposit (Belperio and Freeman, 2004; Belperio et al., 2007; Freeman and Tomkinson, 2010). This suite incorporates the following three units which are locally interbedded within it:
• Algal and micritic laminated and massive dolostone with stylolite-like veins filled with chlorite and hematite. It contains brecciated zones rich in hematite and Cu sulphides, and has carbonaceous dolomite and shale near the base of the unit;
• Poly- and mono-mictic pebble to cobble breccias and coarse-grained clastic to finely laminated argillaceous rock types, dominantly of sedimentary origin, with lesser lithic sandstone beds. The breccias and coarse clastics comprise hematite-rich greywacke-shale breccia with clasts of sandstone, greywacke, siltstone, chert, hematite and basalt in a matrix of crystalline earthy hematite. Rare clasts of retrogressed gneiss are also encountered. Brecciation is strata bound and mainly single stage (Freeman and Tomkinson, 2010), with little evidence of refracturing of previously consolidated breccia. Variable brecciation intensity formed clast- or matrix-supported breccia sheets. Sharp breccia contacts are evidenced by a sudden change of clast size or type. Brecciation predates the hematite-dominant replacement (Schlegel and Heinrich, 2015). This unit is interpreted to represent subaerial, high-energy, clastic sedimentary deposition on an alluvial to fluvial plain immediately below the unconformity separating the Gawler Range Volcanic equivalents and red beds from the stratigraphically underlying shallow marine sedimentary rock package (Bull et al., 2015).
• Andesite with minor dacite, which were probably precursors to the succeeding volcanic sequence;
• Fine grained redbed sandstone with minor red mudstone which occurs as an infaulted block, the stratigraphic position of which is unclear;
• Strongly metasomatised basalt, again occurring as a fault isolated block whose affinity has not been determined;
- Disconformity or unconformity
Footwall volcanic sequence of the 'Neptune Volcanics' (Freeman and Tomkinson, 2010), interpreted to be equivalent to the Gawler Range Volcanic sequence (Belperio et al., 2007; Freeman and Tomkinson, 2010; Bull et al., 2015; Williams et al., 2017). Reflection seismic data reveal that the volcano-sedimentary sequence has a maximum thickness of around 4.5 km and overlies Archaean rocks of the Mulgathing Complex (Harris et al., 2013) ~7 km south of Prominent Hill in a south-verging fold and thrust duplex system. Whilst at Prominent Hill this sequence is overturned, dipping steeply north, it overlies Archaean basement to the south above a contact with a much shallower northerly dip (Harris et al., 2013), inferring that, apart from it being overturned immediately to the south of the Southern Overthrust, the sequence is otherwise relatively shallow dipping. The volcanic rocks of this sequence have geochemical characteristics similar to the Gawler Range Volcanics from elsewhere in the Gawler Craton, including marked enrichment of the most incompatible elements, along with negative Nb-Ta, Zr-Hf and Ti anomalies. Prominent Hill basalts also show strong similarities with a marginal gabbroic facies of the large 1562±14 Ma White Hill Igneous Complex ~5 km to the NW (Williams et al., 2017). The sequence comprises:
• Basalt and basaltic andesite - massive plagioclase porphyritic, amygdaloidal and fragmental facies basalt, basaltic andesite and minor dacite. These coherent footwall volcanic rocks include intercalated fragmental lithologies such as agglomerate, felsic tuff and ignimbrite, and volcanic clast conglomerate (Belperio et al., 2006) that had variably brecciated textures prior to the pervasive hematite replacement and subsequent Cu-(Fe) sulphide mineralisation. The hematite replacement involved extensive alteration to form an assemblage that includes sericite, hematite, chlorite, carbonate and leucoxene altered titanomagnetite.
• Felsic volcanic rocks become much more common several kilometres west of Prominent Hill, with dacite to rhyolite compositions and a distinctive reddish colouration due to hematite dusting of alkali feldspar. This contrasts with the predominantly sericite-chlorite-earthy hematite-leucoxene-carbonate alteration of the basalts and andesites in the deposit area.
- Unconformity or fault
Red-bed sequence with interlayered volcanic rocks - coarse to very coarse grained quartz-rich sandstone and conglomerate with graded and cross-bedding and scours, representing high-energy, oxidized clastic sediments deposited in a subaerial setting, probably an alluvial and/or fluvial plain (Bull et al., 2015). The sedimentary sequence is interrupted by interbedded sericite-chlorite altered amygdaloidal basalt, poorly sorted hematite-chlorite-sericite altered volcanic clast conglomerate and hematitic/sericitic volcanic siltstones. Conglomerates and K feldspar sandstones in the upper section of the red beds are hematite cemented with clasts of vein quartz, ironstone, rhyolite and altered basalt (Freeman and Tomkinson, 2010).
Mafic sills and dykes. The hematite-altered sedimentary host rocks are cut by:
• mostly unbrecciated sericite-chlorite-hematite-silica altered basaltic-andesitic, andesitic and dioritic sills and dykes, some of which have chalcopyrite in the sericite-rich matrix and contain sparse amygdules filled with fluorite, barite and siderite. Chilled margins are not abundant;
• dolerite dykes that post date mineralisation, cut all hanging wall rocks and have been tentatively equated with the ~820 Ma Gairdner Dolerite dyke swarm (Freeman and Tomkinson, 2010).
- Permian unconformity
Phanerozoic cover sequence (after Freeman and Tomkinson, 2010)
• Early Permian Boorthanna Formation - diamictite overlain by rhythmically bedded coarse and fine-grained clastic rocks;
• Lower Cretaceous Cadna-owie Formation - sandstone, siltstone, calcareous sandstone, pebbly sandstone and some feldspathic intraformational conglomerate;
• Late-Early Cretaceous Bulldog Shale - predominantly grey mudstone.

Alteration and Mineralisation
The formation of both hematite-dominated alteration and economic Cu-(Fe) sulphide mineralisation (or at least some) took place during and after rotation of the brecciated host package. This interpretation is based on geopetal structures in the high-grade mineralised breccia, and further supported by the presence of pre-mineral dykes intruding the overturned sedimentary hosts (Schlegel and Heinrich, 2015; Williams et al., 2017). The following alteration, particularly that related to iron oxide, is observed:

• Magnetite skarn developed in the Palaeoproterozoic, probable Wallaroo Group equivalent, metasedimentary rocks of the hanging wall metamorphic sequence. These rocks are altered to a 'skarn' assemblage of massive magnetite with phlogopite-chlorite-pyrite ±chalcopyrite ±actinolite-tremolite ±serpentine ±talc ±scapolite (Freeman and Tomkinson, 2010). Pyrite is abundant, filling fractures and veins within massive magnetite and locally comprising up to 20 vol.% of the 'skarn' assemblage (Belperio and Freeman, 2004), although the Cu grade is typically <0.2%. This assemblage may represent either a single event or be the result of episodic Fe ±K ±Mg alteration, as observed elsewhere in the Mount Woods Domain (e.g., Joe's Dam South East prospect; Schlegel and Heinrich, 2015). In drill core martite (hematite pseudomorphs after magnetite skarn) occurs as zones separated by metre-scale layers of chlorite breccia, granitoid and local fault zones. Magnetite-quartz-calcite-fluorite-barite ±pyrite ±chalcopyrite veins cut all lithologies of the hanging wall metamorphic sequence. Rare calcite-quartz ±hematite ±fluorite ±chalcopyrite veins that have been correlated to the Cu mineralisation in the Main Host Sequence (Belperio et al., 2007), indicating the magnetite skarn predated the hematite-associated economic Cu mineralisation. Rocks composed of the skarn alteration assemblage have been intruded by a dacite porphyry with a crystallisation age of 1586 ±12 Ma (LA-ICP-MS U-Pb dating of zircon; Bowden et al., 2017).

• Hematite breccias in the Main Host Sequence are the result of pervasive hematite replacement of host carbonate and pre-existing siliciclastic breccias. This interpretation is supported by the comparison of textures seen in regional, hematite-barren, calcareous sedimentary rocks and those of hematite breccias from Prominent Hill (Schlegel and Heinrich, 2015). Pervasive hematite can be observed to have variably replaced calcareous and dolomitic lithologies. Although in some cases this replacement has progressed to almost complete obliteration of previous textures, in many cases pre-alteration and pre-mineralistion sedimentary textures and structures are partially retained, e.g., layering, brittle fracturing, angular and rounded clast outlines, boudinage and aligned clast fabrics (Schlegel and Heinrich, 2015). Replacement appears to have commonly increased porosity and probably enhanced permeability of the hematite breccias, compared to their protoliths. There is no hydrothermal magnetite associated with the hematite alteration south of the Hanging-wall Fault, with the only magnetic features reflecting post-mineral dykes (Williams et al., 2017).
The breccias contain a range of clast types, including both fragments of the host rock, as well as red-dusted exotic granitoid clasts. These breccia clasts have a frayed appearance, and in addition to the matrix, may also be partially to totally replaced by hematite. As well as hematite, hydrothermal alteration has produced assemblages including carbonates (siderite, ankerite and dolomite), sericite (phengite), chlorite, fluorite, barite, quartz, fluorapatite, REE minerals (including monazite), uraninite, coffinite and Cu-(Fe) sulphides. Some of these minerals may not be entirely of hydrothermal origin, e.g., chlorite, muscovite and sericite, which are seen to be original constituents of the matrix of unaltered protoliths. Never the less, much of these minerals are formed by hydrothermal alteration and can be seen to be the products of the breakdown of feldspars seen in the unaltered host rocks (Schlegel and Heinrich, 2015). Two end member hematite breccias have been recognised, with transitional varieties developed in brecciated and unbrecciated host rocks. The two hematite breccia types together constitute the Prominent Hill hematite breccia body:
i). hematite-quartz alteration, which is mostly in brecciated rocks, and is characterised by an assemblage of hematite-quartz ±fluorapatite ±barite ±fluorite ±REE minerals ±gold. This alteration style is referred to as 'steely hematite' by Belperio et al. (2007) and Freeman and Tomkinson (2010) and coincides with the gravity anomaly initially drilled that led to the discovery of the deposit. It occurs as a dense, hard rock that overprints earlier mineralised breccia, barren breccia and volcanic rocks. It comprises extremely fine (<10 µm) hematite intricately mixed with cryptocrystalline silica in altered volcanic rocks, whilst in calcareous siliciclastic protoliths it produced cryptocrystalline silica intergrown with hematite. In zones of more intense pervasive alteration, hematitic jasperoid resulted. It contains pervasively hematised and/or silicified angular cherty clasts. Late-stage veins containing barite and specular hematite cut the breccia. Where hematite-quartz is pervasive, both brecciated and unbrecciated host rocks were variably altered, first causing depletion in elements such as Na, K, Mn, Mg and Al and replacement of chlorite and sericite by kaolinite. During this process and subsequently, feldspar, carbonate, amphibole, muscovite and kaolinite were removed and substituted by hematite and/or quartz. Pre-existing breccia textures were progressively destroyed, as indicated by the 'fraying' and irregularly hematite-quartz embayments eroding clast margins. Other characteristics include: a). euhedral to subhedral quartz and hematite crystals projecting into small breccia cavities; b). dull colour changes in semi-altered clasts, and c). a reduction of porosity (Schlegel and Heinrich, 2015). This reduction in porosity, in contrast to the hematite-quartz alteration in the breccia, is interpreted to be due to the more intense hematite alteration and the major addition of silica which progressively occluded permeability and porosity (Schlegel and Heinrich, 2015).
These breccias and the transition to the adjacent hematite-chlorite-sericite-altered rocks host the bulk of the Au (+barite) mineralisation, but are generally poor in Cu-(Fe) sulphides. These enhanced gold concentrations are generally found in a zone some 10 to 20 m wide, typically containing 4 to 6 g/t Au, commonly at the outer edge of the hematite-quartz alteration zones, accompanied by abundant hematite and silica in the breccia matrix. Gold grains, vary in size from <5 to >500 µm, and have rims of copper sulphide and complex symplectic textures with both copper and iron oxides.
ii). hematite-aluminosilicate breccia, containing hematite-chlorite-sericite ±siderite ±ankerite ±kaolinite ±barite ±fluorite ±fluorapatite ±REE minerals [e.g., monazite], uraninite, coffinite ±Cu-sulphide alteration. On the basis of the K/Al ratio in bulk assay data, Schlegel and Heinrich (2015) infer the 'sericite' that is closely associated with the Cu(-Fe) mineralisation is a phengitic muscovite. This 'sericite' has a probable 1585 and 1570 Ma age range (⁴⁹Ar/³⁹Ar; Bowden, 2017). These breccias are generally porous and softer than the hematite-quartz breccia and have textures that vary from being finely layered, to massive, with fine-grained to mm- scale crystalline hematite aggregates, meshes or disk-shaped networks. They are typically dark grey to black, although in contrast, the earthy hematite matrix (see below) on the margin of this breccia is generally fine grained, granular and red to brown (Schlegel and Heinrich, 2015). The porosity and permeability of these breccias which may have contributed to their susceptible to increased fluid penetration is interpreted to be the product of inferred carbonate dissolution and widespread hematite replacement which is likely associated with net mass loss of CaO, Na2O, and variable CO2 and Al2O3 to increased fluid penetration over time (Schlegel and Heinrich, 2015).
Hematite-aluminosilicate breccias are the main host to Cu mineralisation which predominantly occurs within the breccia matrix.
Cu-(Fe) sulphide minerals that are abundant in the breccia matrix include chalcocite, digenite, bornite, idaite and chalcopyrite. These sulphides are variously intergrown with fine- to coarse-grained hematite, sericite, barite, fluorite, carbonate and locally quartz, or replace euhedral to anhedral pyrite (Belperio et al., 2007; Schlegel, 2010). Cu-(Fe) sulphides also infill voids in earlier hematite-quartz altered breccia where the two styles overlap. High Cu grades associated with chalcocite are typically confined to sections of the breccia matrix containing fine- to coarse-grained crystalline hematite and visible sericite (hematite-sericite matrix breccia; see below). Intense hematite-fluorite-barite alteration (hematite-fluorite matrix breccia; see below) is also locally developed in spatial association with chalcopyrite-pyrite mineralisation, occurring in patches adjacent to hematite-only breccia matrix. Discrete accumulations of fluorite and/or barite are intergrown with hematite and Cu-(Fe) sulphides in these patches. Calcite and siderite accompany the hematite-fluorite-barite assemblage and are generally fine grained, filling matrix interstices or are major constituents of veins cutting the hematite breccia (Schlegel and Heinrich, 2015).
Some hematite can be seen to occur as pseudomorphs after feldspars in volcanic clasts, whilst Cu-(Fe) sulphide minerals also replace breccia clasts and unbrecciated host rock, although not homogeneously across the entire breccia or host rock unit. For example, small sub-cm sized clasts are commonly completely replaced by chalcocite or chalcopyrite and fluorite, whereas larger clasts are partially corroded, and fracture controlled alteration of unbrecciated hosts produces pseudobreccias. High grade Cu can occur over short intervals as chalcocite- and bornite-rich replacements of unbrecciated host rock. Where replacement mineralisation impinges on hematite-quartz altered breccia, Cu-(Fe) sulphides are intergrown with barite and together replace breccia clasts, whilst the matrix remains poor in Cu sulphides and barite (Schlegel and Heinrich, 2015).
Variably developed Cu-(Fe) sulphide assemblages, with or without pyrite, are found throughout the Main Host Sequence and upper sections of the footwall volcanic sequence and display complex textures. Those developed in previously hematite-quartz altered breccia usually comprise chalcocite replacing digenite-coated bornite. Rims of Fe-depleted chalcocite and digenite surround bornite relicts in weakly mineralised hematite-quartz altered rocks and in transition zones to hematite-chlorite-sericite altered breccia. However, no other obvious deposit-scale zonation of Cu-(Fe) sulphides or dependence of mineralogy or depth below the palaeoerosion surface has been observed. Paragenetic relationships of mineral precipitation is therefore derived mainly from smaller scale petrographic relationships (Schlegel and Heinrich, 2015). Petrographic observations of the relationships between chalcocite, bornite, chalcopyrite, pyrite and hematite indicate copper, gold, uranium and rare earth elements were introduced contemporaneously with iron oxides. While hematite was introduced in all stages of the system, the relatively rare sulphides are inferred to have developed in the sequence pyrite → chalcopyrite → bornite → idaite → chalcocite (Schlegel and Heinrich, 2015). This progression is consistent with the continued introduction of a Cu bearing oxidised solution to replace Fe in existing sulphides with Cu to advance from pyrite to chalcocite and in doing so, liberate Fe to combine with the excess O to form hematite.
Hematite-aluminosilicate breccias were divided into two sub-grouping by Belperio et al. (2007) and repeated by subsequent authors, reflecting some of the sulphide-gangue associations detailed above. These have been described as follows by Belperio et al. (2007):
- Chalcopyrite-uranium or hematite-fluorite matrix breccia, which are characterised by an assemblage of chalcopyrite-gold ±bornite ±uraninite ±fluorite ±pyrite in a strongly developed breccia that has a purple, dark grey to jet black, porous matrix, with rounded clasts of intrusive, volcanic and sedimentary rocks. This variety commonly displays well-preserved compositional layering and contains late barite-quartz-fluorite veining. Grades typically average ~1.4 wt.% Cu and 0.6 g/t Au, whilst the uranium content locally exceeds 0.5%, commonly associated with high concentrations of fluorite in the matrix and late barite veining. Uranium occurs as sporadic, fine, discrete grains of coffinite and uraninite as inclusion within hematite or chalcopyrite. Rare earth elements cerium and lanthanum are more widespread than uranium, occurring as cerium-bearing apatite, cerium- and lanthanum-bearing monazite, bastnaesite, and a variety of rare earth element oxides throughout the broad iron oxide alteration halo. However, local grains of coffinite also occur in distal crackle-veined earthy breccias. Rare earth element concentrations average about 3000 ppm cerium + lanthanum, whilst silver grades average about 3 g/t, but are generally lower than in chalcocite-bornite zones.
- Chalcocite-bornite or hematite-sericite matrix breccia, which are characterised by an assemblage of chalcocite-gold ±bornite ±covellite ±diginite in a strongly developed breccia that has a bluish-grey, moderately well layered matrix, with sedimentary and/or volcanic clasts that are wispy shaped and sericitised, or rounded and 'rubbly'. These shapes reflect progressive pervasive hematite and/or sericite alteration of coarse clasts of claystone, chert, volcanic rock and sandstone. The matrix comprises platy hematite crystals, granular hematite aggregates, interstitial earthy hematite, and variable quantities of disseminated copper sulphides. Chalcocite and bornite are abundant and relatively fine grained. Grades typically averaging ~2.5 wt.% Cu and 0.6 g/t Au with low to moderate <0.01% U and <400 ppm F compared to the chalcopyrite breccias. Chalcocite is found as both simple grains within the matrix and as complex intergrowths with hematite, whilst bornite occurs as symplectic intergrowths with chalcocite, inferred to suggest coprecipitation. Bornite and chalcocite are also commonly seen to replace earlier formed euhedral pyrite.

Prominent Hill hematite altered sandstone breccia

Prominent Hill hematite altered sandstone breccia

Prominent Hill heteroliithic breccia, hematite altered

Prominent Hill fined grained protolithic sedimentary sandstone breccia

Prominent Hill heterolithic sedimentary breccia

Prominent Hill heterolithic breccia with steely hematite altered matrix

Prominent Hill drill hole DDHURAN, the 'discovery hole'
a - hematite altered brecciated sandstone, at a depth of 242 mm in a 44 m interval that averaged 1.52% Cu, 0.63 g/t Au;
b - hematite altered heterolithic breccia, with sandstone, mafic volcanic and hematite clasts, at a depth of 293 m, in a 35 m interval averaging 4.3% Cu, 0.67 g/t Au;
c - fined grained protolithic sedimentary sandstone breccia with minor hematite in matrix and overprinting hematite bands, at a depth of 310 mm, in a largely un-mineralised interval between 307 and 430 m;
d - heterolithic sedimentary breccia with mafic volcanic and siliceous sedimentary clasts in a hematitic sandy matrix, at a depth of 332 mm, in the same largely un-mineralised interval between 307 and 430 m;
e - heterolithic breccia with sandstone, mafic volcanic and hematitic clasts in a strongly altered to steely hematite matrix, at a depth of 440 m, within an interval between 430 and 466 m that averaged 1.6% Cu, 0.75 g/t Au;
Core from SA Geological Survey Core Library; Images by Mike Porter 2022.

• Earthy hematite alteration and veining - pervasive earthy and granular hematite alteration/flooding and weak brecciation, together with crackle-veined stockwork veining that is predominantly composed of hematite with lesser carbonates, are developed within argillaceous and dolomitic sedimentary rocks peripheral to copper mineralised breccias. The veins are irregular and up to 10 cm thick and occur as crackle and stockwork arrays with a dominant mineralogy of hematite + carbonate + gold. However, in addition to this base assemblage, other mineral packages define four distinct vein types, which in addition to gold, comprise (Schlegel and Heinrich, 2015):
i). Specular hematite-Cu-(Fe) sulphides ±quartz;
ii). hematite-calcite-barite-pyrite ±fluorite ±chalcopyrite ±bornite ±chalcocite ±magnetite;
iii). hematite-siderite-quartz-fluorite-barite chalcopyrite ±calcite, although a variety without barite and calcite, is also locally abundant; and
iv). hematite-barite-bornite-chalcocite.
The same mineral assemblages also fill porous breccia and centimeter-sized cavities in the breccias. In addition, rare paragenetically late fluorite-pyrite-pyrrhotite-magnetite-bearing veins occur locally in the footwall of the Hanging-wall Fault zone. Gold grains within these veins range in size from 5 to 800 µm. Petrological and SEM observations of these gold grains show a persistent association of gold, copper (bornite, chalcocite or chalcopyrite) and hematite, implying a genetic relationship. In particular, gold grains are both rimmed with a copper sulphide, and contain minute copper sulphide inclusions. The rims and inclusions may be either chalcocite or chalcopyrite, and are ubiquitous whether proximal or distal to the high grade copper zones (Schlegel and Heinrich, 2015; Belperio et al., 2007).

A well developed alteration succession can be seen in the volcanic rocks of the Footwall Volcanic Sequence in the structural footwall of the deposit. These rocks were albitised before ore formation. Albite was sequentially overprinted by chlorite, Fe-Mg bearing illite, hematite and sulphides+carbonates. The white mica-hematite alteration intensity increases unevenly towards the orebody, controlled by variations such as the primary porosity, stratigraphic contacts and fracturing. Much of the hematite formed directly as microspecularite with no precursor magnetite. Hematite is commonly associated with minor minerals such as apatite, barite and REE minerals (mainly Ca-REE fluorocarbonates), florencite and REE-only fluorocarbonates that are found in the orebody (Williams et al., 2017). The same authors conclude that the distribution of ore and ore-related alteration at Prominent Hill suggests that fluid flow during mineralisation was strongly influenced by the major discontinuity between the stratigraphic base of the volcanic sequence and the Main Host Sequence.

Prominent Hill Geology

Simplified geological map of the Prominent Hill deposit area below the 90 to 150 m thick Phanerozoic cover overlying a Permian unconformity, and is based on drilling. The main copper-gold bearing clastic and shelf carbonate sequence is located between metamorphic calcsilicate and metasomatic skarn altered rocks to the north, and unmetamorphosed but strongly altered, volcanic rocks to the south. The main copper-gold ore is hosted by the hematite altered polymictic breccias, while the hematite quartz breccias contain variable gold. The location of the principal resource blocks (Main Prominent Hill Cu-Au, Eastern Cu-Au, Western Au and Western Cu) which lie within these breccias are indicated. Modified after: Williams et al. (2017) and Freeman and Tomkinson (2010).

Economic Cu mineralisation at Prominent Hill was introduced comparatively late in the mineral paragenesis (Freeman and Tomkinson, 2010), and comprises relatively abundant Cu sulphides in an oxidised, otherwise low sulphur system. Hematite-dominant alteration continued throughout all of the stages of mineralisation, as evidenced by multiple hematite overgrowths and encrustations. Within the matrix of hematite breccias, early pyrite occurs as euhedral to subhedral crystals, or is intergrown with hematite to form anhedral aggregates (Schlegel et al., 2017). This implies at least two generations of pyrite were deposited, ± weak chalcopyrite, apparently predating much of the Au mineralisation, which in turn, was temporally associated with hematite-quartz alteration and late barite-bearing veins. The Au and barite were succeeded by late hematite-aluminosilicate alteration of brecciated/permeable rocks accompanied by Cu mineralisation that included chalcopyrite [CuFeS2], bornite [Cu5FeS4], idaite [Cu5FeS6], digenite [Cu9S5] and chalcocite [Cu2S] accompanied by barite, with chalcopyrite seen to replace the early pyrite (Belperio et al., 2007; Schlegel, 2010). Bornite coexists with chalcocite and chalcopyrite, but chalcocite and chalcopyrite are not observed together (Belperio et al., 2007). Mineralisation appears to have evolved with the progressive addition of Cu from an early low Cu:S ratio pyrite-chalcopyrite with accompanying hematite to a high Cu:S suite including chalcocite, also with hematite, as recorded by mutually overprinting generations of Cu-(Fe) sulphide assemblages (Schlegel and Heinrich, 2015). This progression is observed in particular in hypogene sediment hosted copper deposits (e.g., the Kupferschiefer ores in Poland; Oszczepalski, 1999). The reaction is chalcopyrite → bornite: 2 CuFeS2 + 3 Cu⁺ + e ↔ Cu5FeS4 + Fe²⁺; and bornite → chalcocite: Cu5FeS4 + 3 Cu⁺ + e ↔ 4 Cu2S + 4 Fe²⁺ (e.g., Macleod and Muir, 1983). Zhao et al. (2014) showed the chalcopyrite to bornite reaction increases in efficiency over 200°C, and in detail, can be achieved by reactions such as 2 CuFeS2 + 3 Cu(HS)₂^- + 3 OH^- ↔ Cu5FeS4 + Fe(OH)2 + 6 HS^- + 0.5 H2O. At lower temperatures (e.g., in porphyry Cu supergene blankets, with continued addition of Cu, coatings and direct replacement of pyrite and chalcopyrite by chalcocite is common, with other secondary Cu-(Fe) sulphide by-products, without increasing S content.

Petrographic observations of the relationships between these sulphides, other minerals and hematite indicate Cu, Au, U and REE were emplaced contemporaneously with iron oxides (Belperio et al., 2007). Oriented drill core containing geopetal markers show economic Cu sulphide mineralisation was deposited during and after the brecciated host strata had been rotated to a steeply dipping attitude, close to the current geometry. This and probable 1585 and 1570 Ma age range (⁴⁹Ar/³⁹Ar; Bowden, 2017) of phengitic muscovite that accompanies economic Cu mineralisation, suggest that the latter occurred after, or late in, the extrusion of the inferred ~1590 Ma lower Gawler Range Volcanic equivalents of the Footwall volcanic sequence (Belperio et al., 2007). Also, the absence of significant Cu mineralisation in the 'magnetite skarn' in the Hanging wall metamorphic sequence, other than regional subeconomic pyrite-(chalcopyrite), suggest it was tectonically juxtaposed after the main stage economic Cu mineralisation at Prominent Hill (Schlegel et al., 2018).

Introduction of Sulphides - On the basis of the observations outlined above, sulphides are interpreted to have been introduced in two stages (Schlegel et al., 2017):
• Stage I, which is characterised by pyrite and minor chalcopyrite in magnetite-altered metasedimentary rocks, part of a pre-ore magmatic hydrothermal sulphide phase responsible for regional magnetite-sulphide alteration and barren to low grade iron oxide accumulations. Deposition of these sulphides was in response to the host rocks interacting with multiple magmatic-hydrothermal fluid pulses that had evolved along a moderately reduced reaction path (Schlegel et al., 2017). Such mineralisation is found throughout the Olympic IOCG Province and Mount Woods Domain, and includes the 'magnetite skarn' above the Hanging-wall Fault at Prominent Hill which has a mineralogy that includes phlogopite + chlorite + pyrite + chalcopyrite + actinolite-tremolite + serpentine + talc + scapolite. At least two early pyrite generations within the matrix of hematite (rather than magnetite) breccias at Prominent Hill, as described above, can be shown to texturally predate high grade Cu mineralisation, and are interpreted to represent this stage I (Schlegel et al., 2017).
• Stage II - petrographic evidence supports Stage I being followed by a late, magmatic derived, oxidised sulphide phase involving multiple pulses of fluid that had first interacted with the oxidised surface environment. Examples of stage II sulphides are observed to have formed by the reaction of stage I pyrite with Cu-bearing fluids to produce Cu sulphides. Petrographic evidence from ore textures, distribution of sulphides and barite and sulphur isotope data are interpreted (Schlegel et al., 2017) to indicate sulphide stage II mineralisation was caused by a combination of two end-member processes, but also a third where sulphides form directly with hematite, namely:
i). magmatic sulphur disproportionation* followed by reduction of magmatic sulphate and precipitation of Cu-(Fe) sulphides. Disproportionation of magmatic sulphur resulted from the mixing of oxygen-rich meteoric waters and acid magmatic hydrothermal fluid at or near the surface: 4 SO2 + 4 H2O + 2 O2 ↔ 4 HSO₄^- + 4 H⁺ This reaction produced sulphuric acid (accompanied by hydrofluoric acid generated from the abundant F within the same fluids) which leached Cu from the enclosing volcanic rocks. The Cu and acid magmatic sulphate were subsequently co-transported to the deposit site where the latter was neutralised by reaction with a reductant containing Fe²⁺ present within, or advected into, the host rock package, as follows: HSO₄^- + 8 Fe²⁺ + 8 H2O ↔ 4 Fe2O3 + H2S + 15 H⁺. This neutralisation of acidity led to the release of disolved Cu that combined with available reduced S in H2S to precipitate Cu-(Fe) sulphides;
ii). isovolumetric replacement of stage I precursor pyrite by Cu-rich sulphides during sulphide stage II. This involves the replacement of previously precipitated magmatic-hydrothermal pyrite by Cu-rich sulphides, with incomplete isotopic equilibration between pre-existing and newly formed sulphides during reaction with an oxidised fluid containing abundant Cu. In the process, Fe²⁺ is expelled as a by-product, as follows: 14 Cu²⁺ + 5 FeS2 + 12 H2O ↔ 7 Cu2S + 5 Fe²⁺ + 3 HSO₄^- + 21 H⁺. Note that this simplified equation is shown to generate chalcocite direct, although observation indicates a progression through chalcopyrite and bornite as detailed previously may have occurred also.
iii). direct formation of hematite and sulphides from a fluid with Cu, limited sulphur as reduced aqueous sulphide, in the presence of high total Fe, in the same fluid, a second fluid or host rock, to form hematite and chalcopyrite: Cu²⁺ + 2 HSO₄^- + 9 Fe²⁺ + 4 H2O ↔ 4 Fe2O3 + CuFeS2 + 8 H⁺. This process is similar to that in Stage I where the fluid has limited S and Cu to produce weak pyrite-chalcopyrite with hematite/magnetite mineralisation. The balance between hematite and magnetite is governed by the equilibrium between the following two reactions 3 H2O + 2 Fe²⁺ ↔ Fe2O3 + H2 + 4 H⁺ and 3 Fe2O3 + H2 ↔ 2 Fe3O4 + H2O. This is influenced by the amount of H2 in the fluids and sinks for H⁺ in acid consuming alteration (Schlegel et al., 2018).
Stage II Cu sulphides, the deposition of which postdated, or overlapped, but outlasted the stage I magnetite- (or hematite)-associated sulphide, are responsible for the bulk of economic mineralisation at Prominent Hill (Schlegel et al., 2017). Stage II mineralisation also postdates the onset of the extrusion of the mafic to intermediate lower Gawler Range Volcanics (Carter et al., 2003; Belperio et al., 2007; Harris et al., 2013) as these rocks are mineralised in the Prominent Hill area.
* NOTE: Disproportionation is a reaction in which a substance is simultaneously oxidised and reduced, giving two different products.

Fluid Inclusions - Applying petrography, microthermometry, and LA-ICPMS microanalysis, Schlegel et al. (2018) were able to characterise pre-, syn- and post-mineral fluid inclusion assemblages and discriminate four fluid end members, as follows:
• Fluid A, interpreted to comprise oxidised surface water containing magmatic solutes. It is the main Stage II ore fluid, and occurs as inclusions in fluorite and barite that are intergrown with Cu sulphides in hematite breccia matrix. It is weakly saline (≤10 wt.% NaCl equiv.) and contains low concentrations of K, Pb, Cs and Fe (600 ppm), but is rich in Cu (~1000 ppm) and U (0.5 to 40 ppm). Its low molar Br/Cl ratio of 0.003 is typical of magmatic fluids, acidic volcanic lake water and evaporated bittern brines, whilst its low Ca and comparatively high K contents, suggests a magmatic-derived origin of the salinity in the inclusion. The Cu and U content are consistent with a fluid that is highly acidic and highly oxidised (e.g., Bastrakov et al., 2010). Its high Al content is also consistent with a highly acidic fluid and the large-scale Al leaching in the core of the Prominent Hill alteration system (Schlegel and Heinrich, 2015). Schlegel et al. (2018) argue the source of this fluid could not have been a playa lake, because such a surface brine would not have been acidic enough to produce the intense hydrothermal alteration, Al leaching and mineral zonation it is interpreted to have produced at Prominent Hill. Schlegel et al. (2018) therefore suggest the solute inventory of the fluid was of magmatic origin, derived through shallow fluid exsolution and degassing of late Gawler Range Volcanics, which were disolved in meteoric water that became enriched in magmatic components. This was followed by complete oxidation of magmatic sulphur via atmospheric oxygen in a volcanic lake environment. The cooling and oxidation of magmatic SO2 in water produced sulphuric acid which aggressively leached Cu and U from eruption rubble in the vent, subaerial lavas and wall rocks of lakes and rivers (Christenson and Wood, 1993). The fluorine-rich nature of the Gawler Range Volcanics magmas (Agangi et al., 2012), suggest the effusive volcanism probably also released HF in gaseous or readily leachable form which would have further contributed to the fluids ability to leach metals. The low Ca concentration of fluid A allowed efficient fluorine transport that eventually became enriched in the deposit by interaction with more Ca-rich brines or host rocks. This fluid is interpreted to have then migrated through oxidised aquifers to the site of the Prominent Hill deposit, where it became the main driver of Stage II copper mineralisation (Schlegel et al., 2018).
An alternate explanation is that the magmatic component of Fluid A was derived from a large, lower crustal mafic to ultramafic magma chamber (as implied by magnetotelluric and deep seismic data; Reid, 2019) rather than having degassed from late Gawler Range Volcanics. As such it may have been derived from fluid collected in the apex of that chamber having segregated from the large volume of magma below, and already be enriched in Cu as well as the other elements detailed above. It rose to shallower depths and mingled with highly oxidised basinal brines in a fault controlled basin filled with red-beds, tuffs and volcano-sedimentary rocks to form the hybrid fluid A, and been modified as described above. If this basin was hydraulically connected to the underlying host sequence, the fluid could have migrated down through oxidised aquifers to the site of the Prominent Hill deposit while scavenging additional Cu and U en route.
• Fluid B, which is a modified magmatic-hydrothermal brine, is found in inclusions hosted by siderite-quartz-fluorite-chalcopyrite and barite-chalcopyrite bearing veins that crosscut the Prominent Hill hematite breccia. Similar fluid inclusion assemblages occur in barren magnetite-quartz veins cutting the weakly mineralised 'magnetite skarn' above the Hanging-wall Fault. It is dominated by NaCl and is hypersaline, the most saline of the four fluids, with a total NaCl+CaCl2 concentration of 36 to 45 wt.% and a low Ca/Na mass ratio of 0.3. It also has the highest concentrations of K, Fe, Mn, Pb and Cs and contains ~70 ppm Cu with a composition that is typical of a moderately reduced magmatic-hydrothermal brine modified by fluid-rock interaction (Audétat et al., 1998, 2008; Ulrich et al., 2001; Samson et al., 2008; Kouzmanov and Pokrovski, 2012). The fluid-rock interaction is interpreted to have involved quartz-feldspar±carbonate bearing rocks, which increased the Ca/Na ratios in the fluid to a moderate degree (Dolejs and Wagner, 2008; Hennings et al., 2017) compared to pristine magmatic fluids. The high temperature of origin of the fluid inclusion assemblage imply that magmatism was active during the formation of the deposit. Schlegel et al. (2018) conclude that the solute components in fluids A and B were both ultimately sourced from magmas with similar composition, but the two are distinct fluids that have evolved along different fluid-flow pathways in the system.
Very similar fluid inclusion assemblages are found in barren magnetite-quartz veins cutting the weakly mineralised magnetite-'skarn' like alteration assemblage in the Hanging wall metamorphic sequence and in the Joe's Dam sub-domain 10 to 12 km to the NW (see images above for location). Although these fluids cannot be temporally linked across the Hanging-wall Fault, neither has produced economic mineralisation and both are tentatively linked as a regional Fluid B. Albite, magnetite and hematite replacement of feldspar and carbonate-rich rocks at Joe's Dam and Joe's Dam South East (Freeman and Tomkinson, 2010) in the vicinity of the fluid-saturating 1562 ±14 Ma (U-Pb zircon; Allen et al., 2016) White Hill Igneous Complex supports a magmatic source of the high temperature Fluid B, modified by exchange of Na for Ca. This interpretation is further supported by the very similar sulphur isotope signature of pyrrhotite and chalcopyrite from magnetite-bearing miarolitic cavities in the marginal gabbro of the intrusion to those of Fluid B, and the presence of magnetite bands in the same gabbro (Schlegel et al., 2017). This implies that late Hiltaba age mafic to ultramafic intrusives may be the source of Fluid B, possibly exsolved and expelled due to mingling with Balta age granites or felsic gneiss country rocks as described above in the Regional Alteration and Mineralisation section.
• Fluid C, a non-magmatic crustal rock-buffered brine that occurs as inclusions in fluorite and barite within bornite-chalcocite bearing hematite-aluminosilicate breccia matrix ores. It is a CaCl2-bearing and NaCl dominated calcic-sodic brine with an intermediate salinity of 16 to 28 wt.% NaCl+CaCl2, an elevated Ca/Na ratio of ~0.6 and high Br/Cl ratios, characteristic of basin brines of residual bittern origin. It carries less Fe than fluid B, but contains ~200 ppm Cu, and likely contributed metals to economic mineralisation. Schlegel et al. (2018) note that the Pb/Cl ratios are similar to those in metal rich basinal brines (Stoffell et al., 2008; Fusswinkel et al., 2013; Wilkinson, 2013; Wagner et al., 2016), supported by the ranges in Ca/K and K/Na mass ratios. Fluid inclusion assemblages similar to fluid C closely overlap data from red bed brines, interpreted to be modified sediment hosted Cu ore fluids (e.g., Kupferschiefer; Fusswinkel et al., 2014). In addition, fluid C data overlap with metalliferous oil field brines from the central Mississippi basin. A component of magmatic salinity is possible in addition to contributions from sedimentary rocks or bittern brines. Fluid inclusion assemblages that approximate end-member fluid C contain ~500 ppm S, which is typical of the sulphate content of basin brines with 30 wt.% total salinity (Hanor, 1994). However, sulphur isotope data provides no evidence of a major sulphate contributions from a sedimentary-evaporitic source, but are readily explained by partial inheritance of S from preexisting pyrite in the host rock sequence (Schlegel et al., 2017).
• Fluid D is a basement brine and is hosted by inclusions in fluorite of late calcite-barite-pyrite-fluorite veins of the ore stage crosscutting the hematite breccia. It is a Ca-dominated and Na-bearing brine with a total salinity range of between 19 and 30 wt.% NaCl+CaCl2, whilst the concentrations of K, Mn, Cs and Pb are similar to those of fluid C. Fluid inclusions that are predominantly composed of fluids C or D have a continuum of Ca/Na ratio, but fairly constant and distinct Cs/Cl and Pb/Cl ratios. The fluid D Ca/Na ratio of is much higher than in typical basin brines. It is interpreted to be a basement brine with characteristically high Ca/Na ratios of ~2, and averages ~100 ppm Cu. This fluid appears to play no significant role in economic Cu-Au mineralisation, as petrographic evidence show that it typically postdates sulphide precipitation in the breccia matrix, although it may have been significant during Cu remobilisation (Schlegel et al., 2018).

Sulphur Isotopes - Schlegel et al. (2017) showed that sulphides from the Prominent Hill deposit and nearby prospects in the surrounding Mount Woods Domain have a wide range in δ³⁴SV-CDT values of between -33.5 and 29.9‰ for Cu-(Fe) sulphides, and a narrower range of 4.3 to 15.8‰ for barite. The Fe sulphides pyrite and pyrrhotite have a narrower range of sulphur isotope compositions, whereas Cu-bearing sulphides, bornite, chalcocite, digenite and idaite on average, have a much wider variation, and more negative δ³⁴SV-CDT values of between -16.2 and -4.6‰.
This wide spread has been interpreted in terms of diverse sulphur sources that include contributions from magmatic, sedimentary, seawater or evaporitic sulphur. In order to test these alternatives, Schlegel et al. (2017) performed a detailed sulphur isotope study of Cu-(Fe) sulphides from Prominent Hill and IOCG prospects (e.g., Joe's Dam) within a 30 km radius which had been influenced by similar regional fluids, particularly Fluid B as described above. This study concluded that the diversity in sulphur isotope composition can be produced by different fluid evolution pathways along reducing or oxidising trajectories, but ultimately derived from a magmatic source that had a δ³⁴SBulk composition of 4.4 ±2‰. This value approximates the magmatic sulphur isotope value determined from pyrrhotite of a Hiltaba Suite-age gabbro in the Mount Woods domain (Bastrakov et al., 2007) similar to that that appears to be the source fluid for mineralisation at Joe's Dam as detailed above (Schlegel et al., 2018). A reduced pathway is indicated for the Stage I sulphide mineralisation related to Fluid B, when intrusion-derived magmatic-hydrothermal fluids produced early pyrite and minor chalcopyrite at Prominent Hill, and iron ±copper sulphides in regional magnetite 'skarns' and in some pervasively altered volcanic rocks of the Gawler Range Volcanics. An oxidised regime was interpreted from related reduced magmatic-hydrothermal fluids vented to the surface, completely oxidised by reaction with atmospheric oxygen to produce sulphate and sulphuric acid with a sulphur isotope composition equal to their magmatic source. These were mixed with oxidised water from the hydrosphere and returned to depth, but contained magmatic solute components, notably sulphate to mix with reduced fluids at depth and react with Stage I sulphide to produce the Stage II mineralisation. Schlegel et al. (2017) argue that their modelling of the sulphur isotope fractionation processes in response to reducing and oxidising pathways demonstrates the entire spectrum of sulphur isotope data from Stage I and Stage II mineralisation can be explained with a single, ultimately magmatic sulphur source. However, some of the early Stage I (or pre-Stage I) pyrite in the host sequence is of sedimentary origin, diagenetic, or partly reworked by early sedimentary brecciation (Allen et al., 2016).

Fluid Mixing and Wallrock Reaction - Schlegel et al. (2018) conclude economic Cu-Au mineralisation at Prominent Hill is the consequence of concurrent fluid mixing and fluid-rock reaction. Mixing of fluids A, B and C is interpreted to have dominated the deposition of hematite, low-sulphidation and low total sulphide Cu-(Fe) mineralisation in the Prominent Hill high-grade orebody, and the main gangue minerals, including fluorite and barite. The same authors interpret fluid A to be the main ore fluid contributing the bulk of the Cu, the relatively small amount of U at Prominent Hill, and oxidised sulphur as sulphuric acid. Fluid B is inferred to have contributed limited sulphur as reduced aqueous sulphide, as well as a high fraction of the total Fe that caused iron oxide deposition. Fluid C contributed some Cu, particularly in peripheral zones with lower ore grades. This fluid-mixing scenario involves three main reservoirs of dominant ore-forming components by precipitation reactions of ore minerals that additionally depend on fluid-rock reactions for acid-neutralising host-rock strata. At the Prominent Hill deposit, the highest copper grades are found at the transition between the most intense hematite-quartz alteration front (Al removal) and hematite-chlorite-sericite altered zones. In these zones, fluid A was neutralised and reduced by reaction with carbonates, aluminosilicates and the Fe²⁺ bearing fluids B and C (and possibly also Fe²⁺ bearing minerals, e.g., chlorite).

Ore Deposit Formation - The following timing and processes in the development of mineralisation and ore at Prominent Hill are proposed, modified after Schlegel et al. (2018):
• Deposition of the Main Host Sequence after ~1750 Ma (possibly nearer 1600 Ma) in an extensional setting, including precursors to the main Gawler Range Volcanic equivalents late in the sequence. This succession unconformably overlies the Mount Woods Domain sedimentary suite, which is interpreted to be equivalent to the 1765 ±27 to 1741 ±1 Ma the Wallaroo Group. The latter was intruded by the ~1691±25 Ma Engenina Adamellite;
• Deposition of the Gawler Range Volcanics equivalents ('Neptune Volcanics') and coeval intrusion of the anorogenic Hiltaba equivalent 1584 ±18 Ma Balta Granites and crust contaminated mantle related 1562 ±14 Ma (U-Pb zircon; Allen et al., 2016) White Hill Igneous Complex and similar 1587±3 Ma (U-Pb zircon; Jagodzinski, 2005) leucogabbro near Peculiar Knob. The underlying mantle lithosphere probably had retained fertility from subduction during the ~1850 Ma Cornian Orogeny when the Donington Suite granitoids were intruded, and during the 1760 and 1739 Ma Kimban Orogeny when the South Australian/Gawler and North Australian (including the Curnamona Province) cratons were amalgamated. The 'Hiltaba age' igneous activity was accompanied by south vergent reverse faulting along structures that were precursors of the Hanging-wall Fault and Southern Overthrust. The resultant uplift to the north led to the deposition of coarse clastics along the fault front. This compressional event was probably the result of transpression and led to basin inversion, and tilting to near vertical of the sequence immediately to the south of the fault. Fluid B was expelled by the mafic to ultramafic White Hill Igneous Complex, possibly the result of mingling with basement felsic gneisses, Engenina Adamellite or Balta granites. This hypersaline fluid, with a high fraction of total Fe²⁺, but limited sulphur as reduced aqueous sulphide, led to the development of Stage I regional iron oxide-alkali (sodic and potassic) alteration, magnetite>pyrite ±weak chalcopyrite mineralisation, and large disseminated to massive magnetite ±pyrite bodies (e.g., Manxman and Joe's Dam).
• Economic mineralisation developed during the late stages in the extrusion and deposition of the Gawler Range Volcanics. Mineralisation at Prominent Hill has been dated between ~1585 and 1570 Ma (⁴⁹Ar/³⁹Ar phengitic muscovite associated with mineralisation; Bowden, 2017). Late volcanism produced comparatively water-poor but hot magmas that released metal-, SO2-, F- and volatile-rich volcanic gases, vapours and fluids into the oxidising environment close to the surface as shallow-venting to subaerial volcanic discharges. These emissions mixed with oxygen-rich meteoric water, most likely in an acidic volcanic lake environment. Such strongly acid (sulphuric and hydrofluoric) fluid is envisaged to have aggressively leached Cu and U from unconsolidated porous tuffs, eruption rubble in the vent, subaerial brecciated and vughy lavas and wall rocks of lakes and rivers (Christenson and Wood, 1993) to form Fluid A. Alternatively, a hot, strongly acidic, F- and Cu-rich magmatic component of Fluid A may have separated from, and accumulated in the apex of a large parental mafic to ultramafic magma chamber (as indicated by magnetotelluric and deep seismic data; Reid, 2019), deep in the crust and followed a transcrustal structure to near surface to mingle with a cool, strongly oxidised brine in a shallow, volcanic fed red-bed basin, to form Fluid A. Whichever the origin, this heavy, cooling, metal-rich, acidic and oxidised, but relatively low salinity Fluid A is envisaged as having migrated downward due to brine reflux, traversing oxidised aquifers, and continuing to leach metals. Fluid circulation is also interpreted to have in part been been driven by topography (Schlegel et al., 2018). The aquifers traversed were presumably porous ignimbrites and tuffs and structures within the Gawler Range Volcanics, as well as the coarse oxidised red-bed units near the base of the volcanic pile, and the coarse clastic rocks below the unconformable base of the volcanic sequence (Bull et al., 2015).
In the first option, to source the amount of Cu in the resource at Prominent Hill of 280 Mt @ 1% Cu, i.e., 2.8 Mt of Cu, from rhyolitic and basaltic Gawler Range Volcanics containing ~50 and ~100 ppm Cu respectively, averaging 75 ppm Cu, it would be necessary to completely leach an ~13.5 km³ volume of aquifer (e.g., a large ignimbrite sheet such as the 30 km³ Taupo Ignimbrite in New Zealand) and efficiently focus the flow of pregnant fluid to and through the Prominent Hill deposit site. To account for loss and surrounding low grade envelopes, etc., a larger volume would need to be leached. If the alternative source is a segregated fluid derived from a very large volume (hundreds of km³) mafic to ultramafic magma chamber at depth, the supply of this amount of copper is far less problematic.
It is suggested that as Fluid A sank further below the surface, its temperature was elevated by heat from underlying igneous activity, thus reducing its density, and promoting upward flow, preferentially via the steep overturned limb of the calcareous Prominent Hill Main Host Sequence which had a permeability structure facilitating efficient fluid infiltration. In this limb, it encountered similarly upward percolating Fluid B which continued to emanate from the White Hill Igneous Complex or similar underlying deeper intrusions, as well as Fluid C resident in pores within the sequence. This resulted in Stage II mineralisation reactions (see the 'Introduction of Sulphides' section above) between the Cu²⁺ and HSO₄^- carried by Fluid A; Fe²⁺ in Fluid B; and the carbonates of the Main Host Sequence wallrocks to precipitate bornite and hematite via sulphate reduction, as follows (Schlegel et al., 2018): 5 Cu²⁺ + 4 HSO₄^- + 37 Fe²⁺ + 40 CaCO3 → 2 Cu5FeS4 + 18 Fe2O3 + 40 Ca²⁺ + 40 CO2 + 2 H2O. Comparable reactions created chalcopyrite and chalcocite, also accompanied by hematite deposition. Similarly the second Stage II reaction, pyrite replacement, occurred as a result of reaction between Cu in Fluid A, and pyrite and carbonate in the host rock to precipitate chalcocite and hematite (and bornite similarly), as follows (Schlegel et al., 2018): 28 Cu²⁺ + 10 FeS2 + 21 CaCO3 + 3 H2O → 14 Cu2S + 10 Fe²⁺ + 6 HSO₄^- + 21 Ca²⁺ + 21 CO2. Limited studies undertaken in sections of the deposit suggest that precipitation of Cu sulphides via pyrite replacement, dominates over that from sulphate reduction (Schlegel et al., 2018).
The reactions between Fluids A and B within each of the two equations in the paragraph above involve the generation of significant H⁺ in the first part of the reaction (e.g., in the sulphate reduction equation): 5 Cu²⁺ + 4 HSO₄^- + 37 Fe²⁺ + 38 H2O → Cu5FeS4 + 18 Fe2O3 + 80 H⁺. This H⁺ is then neutralised by the reaction in the second part of the amalgamated equation which 'decalcifies' the host sequence carbonates and provides space for the deposition of Cu-(Fe) sulphides and hematite from sulphate reduction equation above, i.e., 2 H⁺ + CaCO3 → Ca²⁺ + CO2 + H2O. The excess Ca²⁺ is released to the mixed fluids or combines with the F in Fluid A to precipitate fluorite. The acid neutralisation process leads to replacement of the abundant carbonates by hematite while precipitating Cu-(Fe) sulphides. The molar volumes of the CaCO3 removed and Fe2O3 deposited in its place are similar, and hence the reaction does not occlude porosity or (presumably) permeability, but may allow the focusing of fluid flow and sulphide precipitation, thus further facilitating the alteration reaction. Given that HSO₄^- occurs on opposite sides of the coeval Stage II sulphate reduction and pyrite replacement reaction equations above respectively, little additional acid is required to replace large amounts of carbonate by hematite, and the excess acidity in Fluid A is available to replace further carbonates and feldspars and subsequently other aluminosilicates (e.g., micas), thereby continuing to enhance porosity and permeability while still precipitating hematite (Schlegel et al., 2018).
When the reactive carbonate has been exhausted in the host, or volcanic and/or alumino-silicate containing clastic host rocks dominate, the H⁺ generated by the continuing reaction of Fluids A and B creating Cu-(Fe) sulphide-hematite is neutralised by the alteration of alkali feldspars (e.g., K feldspar) to phyllosilicates (e.g., muscovite) by reactions such as: 3 KAlSi3O8 + 2 H⁺ → KAl3Si3O10(OH)2 + 6 SiO2 + 2 K⁺. As this process continues, aluminosilicate destruction progresses towards the hematite-quartz front to generate kaolinite from muscovite: 2 KAl3Si3O10(OH)2 + 2 H⁺ + 3 H2O → 3 Al2Si2O5(OH)4 + 2 K⁺ and then progresses to release silica and remove Al from kaolinite and combine with the co-precipitated hematite that continues to accumulate and produce, in some instances, almost complete replacement of the host by 'steely' hematite-quartz alteration assemblage: Al2Si2O5(OH)4 + 6 H⁺ → 2 SiO2 + 2 Al³⁺ + 5 H2O.
The 'steely' hematite-quartz alteration zone at Prominent Hill represents the most intense hematite alteration in the mineralised system, reflecting the maximum flux of reacting Fluids A and B, and the locus of the most complete alteration of host rocks. This implies it marks the source direction from which the fluids emanated. The boundary between the strongly oxidised intense hematite-quartz ±barite alteration and the more reduced hematite-aluminosilicate alteration marks an advancing redox front, where the oxidised and extremely acidic ore fluid became reduced and neutralised by ferrous iron from magmatic-hydrothermal or ambient basement fluids and by reaction with chlorite- and carbonate-rich host rocks and formation waters contained within them. Copper is strongly soluble in hot, oxidised (high Eh) and acid (low pH) fluids, and consequently remained in solution in the 'steely' hematite-quartz zone, but was precipitated on passing the redox front into the hematite-aluminosilicate zone where higher Eh and pH conditions prevailed. Cu is virtually absent within the 'steely' hematite-quartz zone, while the highest grade Cu is found in the hematite-aluminosilicate zone adjacent to the redox front where chalcocite is the dominant Cu sulphide. The Cu remaining in the outflowing fluid that had progressively increasing Eh and pH values was precipitated at progressively lower grades and lower Cu:S ratios (i.e., chalcocite → bornite → chalcopyrite-pyrite), to an outer zone of earthy hematite with sulphide veining. This outward decrease in Cu tenor corresponded to a decrease in alteration intensity from kaolin → mica → mica-feldspar → feldspar-mica. As mineralisation progresses, so does the 'steely' hematite-quartz to hematite-aluminosilicate redox front, with Cu being re-disolved from behind and overprinting and reinforcing the grade ahead of the front, as well as the width of the mineralised zone increasing outward from the front. Gold, with differing solubility properties is deposited on the 'steely' hematite-quartz side of the redox front. The location of the deposit was strongly dependent upon a confining corridor of permeable and porous, reactive host rock, namely a pre-existing, carbonate rich sedimentary breccia, into which the fluids could be focused.

Production, Ore Reserves and Mineral Resources

Global reserves and resources at the deposit as of May 2010 were estimated to be:
278.8 Mt @ 0.98% Cu, 0.75 g/t Au, 2.5 g/t Ag.

The declared Ore Reserves and Mineral Resources at Prominent Hill in mid 2008, prior to the commencement of production in 2009, were:
Copper resource - Measured + Indicated + Inferred Resources: 174.20 Mt @ 1.39% Cu, 0.56 g/t Au, 3.4 g/t Ag (0.5% Cu cut-off).
Gold resource - Measured + Indicated + Inferred Resources: 109.2 Mt @ 0.09% Cu, 1.21 g/t Au, 1.0 g/t Ag (0.5 g/t Au cut-off & <0.5% Cu).
TOTAL Resource: 283.4 Mt @ 0.89% Cu, 0.81 g/t Au, 2.48 g/t Ag.
Western Copper resource (additional) - Inferred Resources: 14.5 Mt @ 1.69% Cu, 0.28 g/t Au, 3.7 g/t Ag (0.5% Cu cut-off).

The declared Ore Reserves and Mineral Resources at Prominent Hill in June 2011 were (OZ Minerals, 2012):
Proved + Probable Reserves: 72.3 Mt @ 1.13% Cu, 0.64 g/t Au, 3.03 g/t Ag. (included within resources)
Copper resource - Measured + Indicated + Inferred Resources: 214.9 Mt @ 1.23% Cu, 0.5 g/t Au, 2.8 g/t Ag.
Gold resource - Measured + Indicated + Inferred Resources: 57.8 Mt @ 0.07% Cu, 1.5 g/t Au, 1.1 g/t Ag.

Remaining reserves and resources at Prominent Hill in June 2015 were (OZ Minerals, 2015):
Proved + Probable Reserves: 73 Mt @ 1.0% Cu, 0.6 g/t Au, 2.9 g/t Ag.
Copper-gold resource - Measured + Indicated + Inferred Resources: 152 Mt @ 1.2% Cu, 0.6 g/t Au, 2.8 g/t Ag.
Gold resource - Measured + Indicated + Inferred Resources: 27 Mt @ 0.1% Cu, 1.3 g/t Au, 1.5 g/t Ag.

Remaining Ore Reserves and Mineral Resources at Prominent Hill at 30 June 2017 were (OZ Minerals, 2017):
Ore Reserves
Open Pit - Copper
Proved + Probable Ore Reserves: 8 Mt @ 1.0% Cu, 0.6 g/t Au, 3 g/t Ag.
Underground - Copper
Proved + Probable Ore Reserves: 39 Mt @ 1.4% Cu, 0.6 g/t Au, 3 g/t Ag.
Surface stockpiles - Copper
Proved Ore Reserves: 12 Mt @ 0.8% Cu, 0.4 g/t Au, 2 g/t Ag.
Surface stockpiles - Gold
Proved Ore Reserves: 15 Mt @ 0.1% Cu, 0.8 g/t Au, 2 g/t Ag.
TOTAL Proved + Probable Ore Reserves Copper + Gold: 74 Mt @ 1.0% Cu, 0.6 g/t Au, 3 g/t Ag.
Mineral Resources - Copper (includes Ore Reserves)
Open Pit (0.25% Cu cut-off)
Measured + Indicated + Inferred Mineral Resources: 8 Mt @ 1.0% Cu, 0.6 g/t Au, 3 g/t Ag.
Underground - (AUD 57 NSR envelope cut-off)
Measured + Indicated + Inferred Mineral Resources: 120 Mt @ 1.2% Cu, 0.6 g/t Au, 3 g/t Ag.
Surface stockpiles
Measured Mineral Resources: 12 Mt @ 0.8% Cu, 0.4 g/t Au, 2 g/t Ag.
TOTAL Measured + Indicated + Inferred Mineral Resources - Copper: 140 Mt @ 1.2% Cu, 0.5 g/t Au, 3 g/t Ag.
Mineral Resource - Gold (includes Ore Reserves)
Open Pit (0.25% Cu equiv. cut-off)
Indicated + Inferred Mineral Resources: <0.5 Mt @ 1.0% Cu, 0.9 g/t Au, 1 g/t Ag.
Underground - (AUD 57 NSR envelope cut-off)
Indicated + Inferred Mineral Resources: 7 Mt @ 0% Cu, 2.4 g/t Au, 1 g/t Ag.
Surface stockpiles
Measured Mineral Resources: 15 Mt @ 0.1% Cu, 0.8 g/t Au, 2 g/t Ag.
TOTAL Measured + Indicated + Inferred Mineral Resources - Gold: 23 Mt @ 0.1% Cu, 1.3 g/t Au, 2 g/t Ag.
TOTAL Measured + Indicated + Inferred Mineral Resources - Copper + Gold: 163 Mt @ 1.04% Cu, 0.6 g/t Au, 3 g/t Ag.

Production in 2016-17 was 112 008 t of Cu metal and 3.94 t Au

For more detail see the references listed below. The title to the deposit is held by Oxiana Limited who are undertaking a feasibility study.

Return to top

Cloncurry Workshop ............. Thursday 24 November, 2005.

A field, drill core and classroom workshop was held in the Cloncurry Terrane of north-west Queensland, for the full day on Thursday 24 and on Friday 25 before and after the Ernest Henry visit. It was led by Dr Patrick Williams of James Cook University (Townsville, Australia), an internationally known expert on IOCG deposits, particularly those in the Cloncurry Terrane. He addressed the setting and characteristics of these deposits in an intitial classroom session, before going into the field to study IOCG mineralisation, alteration (local and regional) and lithologies both in outcrop and in drill core from the E1, Monakoff and Great Australia IOCG deposits of Exco Resources. On Friday 25th, a field visit was made to the regionally altered exposures and outcropping magnetite mineralisation at Mount Fort Constantine near Ernest Henry, while core was examined from the Universal Resources Little Eva deposit and presentation received on its geology and mineralisation.

Return to top

Ernest Henry .................... Friday 25 November, 2005.

The Ernest Henry IOCG style Cu-Au deposit is located 35 km NE of Cloncurry, 150 km east of Mt Isa and 750 km west of Townsville in north-west Queensland (#Location: 20° 26' 40"S, 140° 42' 21"E).

The Ernest Henry deposit lies within the Cloncurry-Selwyn zone of the Eastern Fold Belt of the Mount Isa Inlier in north-west Queensland. The Cloncurry-Selwyn zone represents the eastern exposed margin of the Mount Isa Inlier, bounded to the west by the broad, up to 7 km wide, Cloncurry Fault Zone. This fault zone is the result of multiple stages of a long-lived deformational history, and includes the transcrustal Cloncurry Lineament, the steep Cloncurry Fault and the WNW vergent Cloncurry Overthrust (see Austin and Blenkinsop, 2010). The deposit is hosted by Palaeoproterozoic 1760 to 1660 Ma silici-clastic metasedimentary and metavolcanic rocks that were deposited during periods of ensialic rifting and are part of the greater Eastern Succession of the Eastern Fold Belt. Regionally, the host sequence is intruded by the 1540 to 1500 Ma, A-type Williams and Naraku granite batholiths that are interpreted to be temporally coeval with IOCG mineralisation in the Eastern Fold Belt. Ernest Henry also lies ~10 km to the west of the curvilinear, concave to the east, and west dipping Gidyea Suture Zone. This suture forms the eastern margin of the Mount Isa Inlier, separating it from the older Archaean and/or Palaeoproterozoic Numil Seismic Province to the east. Subduction that culminated in this suture is interpreted to be related to the 1900 to 1870 Ma Barramundi Orogenic event that is widely represented in the North Australian Craton. As such, at the time of deposition, Ernest Henry may have overlain a legacy fertilised sub-crustal lithospheric mantle.

For more detail on the regional geological, structural and metallogenic setting, as well a geological summary map, see the Cloncurry Iron Oxide Copper Gold Province record.

In summary, the Eastern Succession of the Mount Isa Inlier comprises a poly-deformed Palaeo- and Mesoproterozoic volcano-sedimentary succession which is largely composed of evaporite-rich Cover Sequence 2 and silici-clastic-rich Cover Sequence 3 rocks (CS2 and 3). CS2 and 3 were deposited between 1790 and 1690 Ma and from 1680 to 1610 Ma respectively. To the west, these sequences overlie an older crystalline basement and a core of predominantly Cover Sequence 1 felsic volcanic and related intrusive rocks that correspond to the 1870 to 1850 Ma Barramundi Orogeny of northern Australia. Basement is not exposed in the Cloncurry district. Both CS2 and 3 were deposited in intracontinental rift settings, although the relationship between some parts of the sequence is obscured by the deformation history. Both sequences were also accompanied by the emplacement of various intrusive and volcanic rocks.

The first significant deformation to affect CS2 (but not CS3) was the 1750 to 1735 Ma Wonga extensional event. CS2 was extensively intruded by the 1750 to 1730 Ma Wonga Granite to the west, while the coeval Mount Fort Constantine volcanics are found to the NE. Minor tonalites, granitoids and diorite emplaced between CS2 and 3 have been dated at 1686 to 1660 Ma (including the Ernest Henry Diorite).

Thin skinned deformation of the ~1600 to 1520 Ma Isan Orogeny terminated deposition of Cover Sequence 3, and resulted in gross eastward tectonic transport, interleaving of major lithostratigraphic units, and a dominant north-south tectonic grain. This deformation has been divided into: a D1 event, which involved overall north-south compression, and is characterised by large-scale thrusts and isoclinal folds, thrust reactivation of large, km-scale, basin bounding extensional faults with CS3 rocks thrust over CS2, resulting in overturned limbs and a penetrative rock mass foliation; a D2 event, involving horizontal east-west compression producing major north-south upright to isoclinal folding of CS2 and 3 rocks, and a penetrative cleavage, which peaked at 1595 to 1580 Ma with a regional greenschist to upper amphibolite facies metamorphism and the development of anatectic pegmatites; and a D3 event which includes NW- and NE-trending brittle-ductile corridors of faulting, kinking and folding with steep plunges to the NW and NE, and dominantly north-south trending shear and fault zones and associated breccia formation.

Both CS2 and 3 were intruded by the voluminous Williams and Naraku granite batholiths at 1540 to 1500 Ma (including the 1530 Ma Mt Margaret Granite immediately to the east of the E1 deposits; Marshall and Oliver, 2007; Page and Sun, 1998). These represent the youngest felsic intrusions in the inlier, and have an outcrop exposure of >1500 km². They were emplaced in an intracratonic environment, and have a pre-, syn- and post-D3 timing, and are largely composed of alkaline to sub-alkaline, K-rich, A-type, magnetite-bearing granitoids. They range from diorite to syenogranite in composition and are typically more oxidised than similar older (~1670 Ma) granitoids in the Western Fold Belt of the Mount Isa Inlier. Sodic intrusions of similar age are rare. Ernest Henry geological pit plan

A regionally extensive Na-Ca hydrothermal system in the Cloncurry district (>1000 km²) affected all rock types, especially the resultant calc-silicate-rich lithologies of cover sequence 2. This alteration appears to have been formed by multiple periods of hydrothermal activity that locally overlapped and is most intense in breccia zones along large structural conduits and within calc-silicate-rich units. The bulk of the sodic-calcic alteration, dominantly regional albite and scapolite, was associated with fluids that were initially mostly sedimentary formation waters with lesser magmatic components, prior to and during peak metamorphism (Kendrick et al., 2008; Oliver et al., 2008; Baker et al., 2008). Subsequent more structurally controlled albite-actinolite-magnetite-titanite±clinopyroxene assemblages, were synchronous with major granite (e.g., Williams-Naraku batholiths) emplacement (Baker et al., 2008), involving a larger magmatic fluid component, and coincided with formation of the majority, but not all of the significant oxide Cu-Au deposits. These deposits may have some stratigraphic control, but are usually associated with brittle and brittle-ductile shear and fault structures which acted as conduits for the transport of high temperature (300 to 500°C) saline fluids into the host rocks (Williams, 1998).

The Ernest Henry deposit is concealed by 35 to 60 m of extensive Phanerozoic cover and does not outcrop. While the exact stratigraphic position of the host rocks is not known, they have been tentatively correlated with the 1730 ±10 Ma Mount Fort Constantine Meta-volcanics towards the top of Cover Sequence 2. The Mount Fort Constantine metavolcanics comprise dacite and andesite with subordinate metabasalts and calc-silicate metasedimentary rocks. The only other outcrop in the district is the 1480 Ma Mount Margaret granite some 12 km to the east. Within Cover Sequence 2, volcanism is common between 1790 and 1780 Ma, and 1760 to 1720 Ma, with later 1540 to 1450 Ma granitoids.

Ernest Henry Longitudinal section

Within the immediate orebody area the principle lithologies encountered are: i). altered plagioclase phyric andesitic volcanic/hypabyssal rocks (ca 1740 Ma) which host the orebody where they are brecciated; ii). various siliciclastic, calc-silicate-rich and graphitic metasedimentary rocks that occur as <10 m thick intercalations within the metavolcanic rocks; and, iii). medium-grained metadiorite (ca 1660 Ma).

Structural analysis suggests that ore deposition accompanied reverse-fault movement between two northeast trending bounding shear zones and formed a pipe-like zone of dilation in the K-feldspathised metavolcanic rocks. The breccia pipe, plunges at ~45° to the SSESSE, nested between the ductile shear zones (Rusk et al., 2010). The orientation of this dilational zone is consistent with the shape and dip of the Ernest Henry ore breccia.

Ernest Henry magnetite-chalcopyrite ore

Images: Upper right - High grade Ernest Henry magnetite-chalcopyrite-bornite breccia mineralisation.
Lower right - clast in breccia (2 x 1.5 cm clast in centre of image), surrounded, and being replaced by matrix of magnetite, sulphides and quartz. Images by Mike Porter, 2013.

Four stages of alteration are recognised at Ernest Henry:
i). Regional pre-ore Na-Ca alteration, occurring mainly as albitic plagioclase-, magnetite-, clinopyroxene- and amphibole-rich veining and fault-related breccia-fill.
ii). Pre-mineralisation potassic-(manganese-barium) alteration which only contains minor sulphides, and is typified by multiple stages of K feldspar-, biotite-, amphibole-, magnetite-, garnet- and carbonate-bearing veins, and by fault-related breccia and alteration.
iii). Mineralisation associated alteration, characterised by K feldspar veining and alteration, although this may in part be hematite dusted feldspar. K feldspar alteration is most intense in the vicinity of copper-gold mineralisation, but forms a halo extending from several hundred metres up to 2 km beyond the ore body (Mark et al., 2006), although this outer halo may represent part of pre-ore regional alteration zone. Mineralisation is divided into two main stages, characterised by similar mineral assemblages. The first stage of economic Cu-Au mineralisation was the main ore-forming event, associated with a matrix-supported hydrothermal breccia that is enveloped by crackle veined K feldspar altered meta-volcanic rocks. The second stage of mineralisation occurs as a network of veins cutting earlier infill-supported ore-breccias, and contains a largely identical mineralogy to earlier stage. The ore-bearing assemblage dominantly comprises magnetite, pyrite, chalcopyrite, carbonate and quartz, with lesser apatite, barite, titanite, actinolite, biotite and fluorite. In the upper levels of the deposit, the bulk of the ore is present as hypogene chalcopyrite infilling between K feldspar-altered breccia clasts, while at greater depths, it both infills between, and replaces clasts. Electrum and native gold are closely associated with pyrite and chalcopyrite (Foster et al., 2007).
iv). Post-ore, volumetrically minor, multiple stage calcite-dolomite- and/or quartz-rich veining and alteration which lacks magnetite, and only carries a little gold. Deeper in the deposit, breccias include rounded clasts of previously mineralised breccias containing magnetite, pyrite and chalcopyrite, indicating multiple superimposed brecciation events (Rusk et al., 2010).

Ernest Henry magnetite-hematite-feldspar breccia

Ernest Henry magnetite-hematite-feldspar breccia

Image right - Ernest Henry magnetite and K feldspar-hematite breccia with chalcopyrite mineralisation. Polished sample. Sample collected by Richard Lilley; Image by Mike Porter, 2021.

Rusk et al. (2010) interpret the data from Ernest Henry to be consistent with the following genetic trend:
i). Rapid devolatilisation (of possibly both chloride-rich brines and CO2-rich fluids) within the source magma chamber;
ii). Fluid over-pressuring in the roof of the magma chamber as a result of volatile exsolution and vapour expansion, assisted by a seal created by magma solidification, sodic-calcic alteration and/or contact metamorphism in the carapace of the igneous complex;
iii). Possible leakage of over-pressured magmatic fluid along structures controlling the location of the later breccia pipe, producing a pre-ore potassic alteration halo;
iv). The eventual failure of the seal and sudden release of fluid pressure, resulting in a high-energy fluid flow event driving brecciation and upward transported and milled clasts. The resultant breccia mass permitted the mixing and/or subsequent ingress of basinal brines circulating within fractured rocks several kilometres above the magma chamber. Fluid mixing, rapid depressurisation and resultant cooling led to ore precipitation within the matrix porosity between breccia clasts at the top of the orebody, where, as the fluid flow, temperature and pressure declined the breccia was sealed;
v). At depth, closer to the heat source, the temperature and pressure gradient degraded more slowly, allowing for fluid-rock reaction to be more protracted, such that prolonged chemical interaction between K feldspar-rich host rocks and ore fluids led to replacement style mineralisation within clasts, with the same mineral assemblage as observed in the shallower parts of the deposit.
vi). At the deepest levels, repetition of the cycle may have resulted in the release of a new pulse of fluids which brecciated and tapped earlier formed magnetite-chalcopyrite rich rocks, telescoping mineralised clasts upwards into the orebody along narrow channels, thereby upgrading ore.

Ernest Henry mineralised brecciated K feldspar

Ernest Henry mineralised brecciated K feldspar

Image right - Ernest Henry brecciated hematite dusted K feldspar with chalcopyrite and magnetite veining. Polished sample. Sample collected by Richard Lilley; Image by Mike Porter, 2021.

The brecciated volcanic mass that hosts the ore forms a plunging elongate body, some 250 m thick, 300 m average length and extending at least 1000 m down plunge to the SSE. The breccia ranges from the unbrecciated volcanics, to crackle fracture veining to clast supported and matrix supported breccia to total clast digestion (massive matrix). The breccias typically contain 5-20 mm subrounded to rounded meta-volcanic and rare biotite altered meta-sedimentary clasts. The matrix is largely composed of magnetite, calcite, pyrite, biotite, chalcopyrite, K feldspar titanite and quartz. Accessory minerals include garnet, barite, molybdenite, fluorite, amphibole, apatite, monazite, arsenopyrite, a LREE fluorcarbonate, galena, cobaltite, sphalerite, scheelite, uraninite and tourmaline. The bulk of the economic mineralisation is restricted to breccia zones with more than 10% matrix.

The total reserve + resource prior to the commencement of mining in 1998 was 166 Mt @ 1.1% Cu, 0.54 g/t Au.
As of June 2003 the remaining resource totalled 117.9 Mt @ 1.13% Cu, 0.52 g/t Au.
At 30 June 2006, the reserves and resources were (Xstrata, 2007):
Open cut proved reserves - 41 Mt @ 0.9% Cu, 0.5 g/t Au + probable reserves of 20 Mt @ 0.8% Cu, 0.4 g/t Au,
Open cut measured + indicated resources were the same as, and included the proved and probable reserves,
Open cut inferred resources - 1 Mt @ 0.4% Cu, 0.2 g/t Au,
Underground indicated resources - 21 Mt @ 1.5% Cu, 0.7 g/t Au + inferred resources of 23 Mt @ 1.4% Cu, 0.7 g/t Au,

Remaining Mineral Resources and Ore Reserves as at December, 2011 (Xstrata, 2012) were:
Open pit
Total resource and reserve - depleted during 2011 from 17 Mt @ 1.0% Cu, 0.5 g/t Au, 23% magnetite at December 31, 2010
Underground:
Measured resource - 4 Mt @ 1.3% Cu, 0.7 g/t Au, 32% magnetite
Indicated resource - 71 Mt @ 1.3% Cu, 0.7 g/t Au, 28% magnetite
Inferred resource - 13 Mt @ 1.2% Cu, 0.6 g/t Au, 26% magnetite
Total resource - 88 Mt @ 1.3% Cu, 0.7 g/t Au, 28% magnetite
Total ore reserve (all probable) - 74 Mt @ 0.95% Cu, 0.5 g/t Au, 23% magnetite.

Remaining Mineral Resources and Ore Reserves as at 31 December 2020 were (Glencore Resources and Reserves Report, 2020):
Underground
Measured resource - 4.7 Mt @ 0.93% Cu, 0.51 g/t Au
Indicated resource - 55.2 Mt @ 1.16% Cu, 0.61 g/t Au
Inferred resource - 15.5 Mt @ 1.17% Cu, 0.62 g/t Au, 26%
Total resource - 75.4 Mt @ 1.15% Cu, 0.61 g/t Au
Total ore reserve (Proved + Probable) - 38.5 Mt @ 0.95% Cu, 0.50 g/t Au.
NOTE: Resources include Reserves.

The operation is controlled by Ernest Henry Mining Pty Ltd, a subsidiary of Glencore plc. In October 2016 Evolution Mining purchased an economic interest in the copper and gold production from Ernest Henry Mining. In November, 2021, Glencore and Evolution Mining Limited entered into a binding agreement for the sale and purchase of Glencore’s 100% interest in Ernest Henry Mining Pty Ltd.

Return to top

Travelling from Cloncurry to Brisbane .................... Saturday 26 November, 2005, arriving by 1:30 pm.

The summaries above were prepared by T M (Mike) Porter from a wide range of sources, both published and un-published. Most of these sources are listed on the "Tour Literature Collection" available from the IOCG 05 Tour options page.