Search

Significant growth in the size of gold Resources on many mines of the Yilgarn Craton has occurred well after the deposit was discovered and after mining had commenced. Examples include the Kalgoorlie and Norseman goldfields found more than a century ago and the Jundee and Sunrise goldfields discovered in the 1990s. The initial discovery event is referred to as greenfields exploration, and the later additions as brownfields exploration. For the Yilgarn Craton, annual Resource additions have significantly exceeded mine depletion since 1980. Brownfields exploration enjoys economic benefits not available to greenfields exploration such as reduced infrastructure and regulatory costs. In Australia, criticism of the exploration and mining industry for failing to find deposits is misplaced.

Field-based studies can place constraints on the shape of granite intrusions with the results being complementary to the less direct approach using geophysics. The Tallarook plateau in central Victoria provides the necessary outcrop and topographic relief to allow mapping of the outer boundary of the granite in detail, and to infer a thin sheet-like form for an intrusion that was previously interpreted as a vertically extensive cylindrical body. The boundary of the Tallarook granite against hornfels is box-shaped in plan with sides of 5–10 km length. These sides are quite irregular on the scale of 100s of metres with many near-vertical and near-horizontal intervals that define a step-and-stair geometry. There are also intervals with moderate outward dip, and fewer inward-dipping intervals. Various parts of the exposed contact are inferred to be the roof of a granite sheet, sides, and the underside towards a floor. The near-horizontal intervals that delineate the roof of the granite are from 200 to...

This is a book I would have appreciated throughout my career in the mineral industry and in university teaching. It is broad in what it covers but also detailed in the amount of information it has ...

Alteration is recorded throughout much of the 3 km thick Upper Witwatersrand succession in South Africa, and extends for 300 km around the basin and 50 km into the basin. The evidence for this alteration comes from petrography, mineralogy, geochemistry, rock fabrics and stratigraphy. The alteration, conceived as a giant halo enveloping each major goldfield, requires a major hydrothermal event focused on the auriferous Upper Wits (Central Rand Group). The scale and nature of the alteration, and its relationship to gold distribution on various scales, strongly support a hydrothermal model for gold deposition. Such a model can account for the 50,000 t of gold mined from the Witwatersrand goldfields. The model works within the context of the major advances in understanding of the geological framework of the Witwatersrand Basin including the recognition of the importance of post-burial processes. It is highly significant that the age of Witwatersrand sedimentation is much older than 2700 My, and thus the host rocks for gold were deposited at least 100 My before the major global Archean greenstone gold event. The shale units from 40 major mines covering the seven goldfields are examined to establish the nature and distribution of the alteration, and to make the connection between the fluid infiltration that caused the alteration, and the introduction of gold into the Witwatersrand Basin. In the hydrothermal model described here, an auriferous low salinity fluid is produced at about 500 °C during the metamorphism of mafic material that was below and adjacent to the Witwatersrand Basin. Large shear zones in the footwall of goldfields are the inferred fluid pathways into the Upper Wits succession, and then unconformities, smaller shear zones, veins and the reef packages guide local fluid flow. In this model, gold precipitation occurs in response to the reaction of the fluid with carbon- and iron-bearing rocks that are abundant in the reef packages immediately overlying unconformity surfaces. Gold precipitation occurred at about 350 °C and 10 km depth with the implication of a high geothermal gradient approaching 35 °C per km. The alteration of the shales, but lack of abundant gold in them, is predicted from the paucity of appropriate rocks or minerals in the shale units to precipitate gold, but fluid infiltration still changed their composition fundamentally. The source of the Witwatersrand gold, its present distribution and the alteration are explained within this hydrothermal model, with similarities to the models invoked for other major gold provinces. It is worth noting that quartz-pebble conglomerate is a poor term for this gold type and has almost certainly limited exploration opportunities globally.

Well-considered schemes that classify ore deposits can be effective contributors to exploration success, understanding of deposit genesis, mining geology, engineering, metallurgy, property valuations and stock market analysis. An effective classification scheme needs to be easy to apply without ambiguity and is much more powerful if based on sound scientific principles. For deposits with economic gold, a useful classification recognises gold-only and gold-plus deposits based upon whether they also have economic base metals. The scheme is easy to apply and in most cases unambiguous for specific deposits and whole provinces. The subdivision has a sound scientific basis that involves the chemistry of gold in relation to metal complexing, salinity of ore-forming fluids and redox state. Forward modelling from chemical principles can be integrated with inverse modelling from observations to add confidence in this classification and thus be used to create exploration opportunities. The Jiaodong gold province of eastern China comprises deposits mostly in granitic rocks with economic gold but low levels of base metals; the inferred ore fluids are of low to moderate salinity. From their gold-only character, several other aspects of their development can be inferred, though the tectonic and metamorphic history remains unresolved. The granitic host rocks appear to reflect a chemically and mechanically favourable site for deposition, not a source for the gold or the ore fluid neither of which have been established yet.

Discovery of a goldfield is a rare and difficult event, which contrasts with the normal outcome of exploration that is non-discovery. As there are many gold provinces globally and numerous discoveries annually, the recent and historical records of success are a great source of ideas to learn from and to enhance the rate of discovery.Over the last two decades, important gold discoveries have been made in well-established gold provinces such as Australia, Canada and USA, regions with a long history of production like Ghana, and some further countries such as China, Mali, Tanzania and Peru that received lesser attention during the twentieth century. In contrast to the dominance of gold-only deposits throughout the last millennium of gold production, a significant proportion of recent discoveries are gold-plus, especially Cu–Au.There have been some golden periods of exploration success in various parts of the world and as well as providing inspiration for exploration, they are sources of learning. Exa...

The opening of the Yandal gold province during the 1990s provides a detailed case history of repeated exploration successes. First highlighted in 1992 with the discovery of the Bronzewing gold depo...

Goldfields extend for 300 km around the margin of the Archaean Witwatersrand Basin of South Africa associated with regional greenschist facies metamorphism and deformation. Metamorphic mineral assemblages involving pyrophyllite-chloritoid are associated with gold in all goldfields, reflecting low pressure and 300 to 400°C conditions, and indicating high geothermal gradients. The origin of Witwatersrand gold can be explained by metamorphic devolatilisation to generate auriferous fluids beneath and outside the Witwatersrand Basin, followed by passage of these fluids along large structures and into the Basin. In this model, generation of the fluid is a consequence of the transition to amphibolite facies of mafic rocks such as Archaean greenstone belts, as in . In a grain-by-grain scale devolatilisation process, the lower temperature minerals have released their H2O, CO2 and H2S to form a metamorphic fluid which at the time of its formation has already dissolved gold from these unenriched rocks (say a...

After a century of virtual neglect, exploration in the Yandal greenstone belt of the Yilgarn Craton of Western Australia has yielded resources of 12 Moz Au during the 1990s. Success has come from a combination of conceptual geological models, surface prospecting, understanding the weathering environment, and systematic drilling. The Archaean Yandal greenstone belt comprises a lowermost banded iron formation, extensive basalt and dolerite sills, ultramafic rocks, intermediate to felsic volcanic rocks, and variable clastic sedimentary rocks. Early shear zones trend NNW and form the greenstone belt margins, or trend N–S within the belt. Later brittle cross-faults are critical in gold localization. Gold resources and past production at major deposits include Bronzewing (4 Moz Au), Jundee (5 Moz) Mt.␣McClure (1 Moz) and Darlot (3␣Moz, some of which was produced before 1990). All major deposits are hosted by Fe-rich mafic rocks, and mineralization displays a combination of different orientations and morphologies. Quartz veins are surrounded by broad carbonate alteration with proximal K-mica and Fe-sulphides. The recognition of a critical role for the late brittle structures in localizing gold implicates mid-crustal processes within the greenstone belt for fluid generation, and with the host rock control, supports the model in which fluid was derived by metamorphic devolatilization.

The Palaeozoic succession of Victoria represents a major world gold province with a total production of 2500 t of gold (i.e. 78 million oz). On a global scale, central Victoria represents one of the most gold mineralized areas outside the Witwatersrand of South Africa, and remains the prime example of a ‘slate belt’ gold province (also known as ‘turbidite-hosted’, or ‘shale-greywacke’ gold province). Production came from quartz veins (1000 t), modern placers (1200 t) and palaeoplacers (300 t). Up to 7000 mines produced gold from quartz veins; however, the 168 mines of this type which produced over 1 t Au contributed 68% of the primary gold production. Twelve goldfields have each produced 30 t Au (approximately one million ounces) from all sources, with Bendigo (697 t), Ballarat (408 t) and Castlemaine (173 t) being the largest: these twelve goldfields have contributed 71% of the combined primary and secondary production. Gold mining commenced in 1851, declined dramatically around 1914, and has not increased significantly since. The Palaeozoic succession of Victoria is part of an extensive fold belt along much of eastern Australia (i.e. the Tasman Orogenic Belt). In Victoria, the succession is dominated by Cambrian to Early Devonian clastic metasedimentary rocks that have undergone major upright deformation and low-grade metamorphism, culminating in the Middle Devonian (Tabberabberan) Orogeny. North-south inliers of Cambrian metavolcanic- and metasedimentary rocks associated with regional fault zones divide the province into several geologically distinct zones. In central Victoria, there are particularly well-developed acid volcanic complexes of Late Devonian age, and synchronous peraluminous granites which represent high-temperature crustal melting. Tertiary to Quaternary basalts have covered some of the Palaeozoic gold deposits, and have also covered several rich gold placer deposits of Cainozoic age. Larger goldfields, including Bendigo and Ballarat, occur in Ordovician slate and greywacke sequences, but a wide range of metamorphosed sedimentary and igneous rock types are host to goldfields. Most, if not all, primary deposits demonstrate structural control, although the controlling structural features differ between deposits. Wallrock alteration is strongly influenced by host rock: in metasedimentary rocks, it is limited and subtle, but it is more pervasive in mafic and felsic igneous rocks. Carbonates, muscovite and pyrite are the most widespread alteration minerals in these rocks, representing CO2, K and S addition. Enrichment of As within gold deposits is common, with Sb important in some small deposits, whereas base metals are only abundant locally, and Bi, W, Mo and Te show strong spatial association with granites and are rarely associated with economic gold deposits. Granites are an important component of the Palaeozoic succession in Victoria. No large deposits are within granites, but a few goldfields, including Maldon (65 t Au), are within contact aureoles of either S or I type granites of differing degrees of fractionation and silica content. A single-stage, but protracted and possibly diachronous, period of gold introduction near the time of the Middle Devonian (Tabberabberan) deformation can explain geological relationships at many gold deposits. The possibility that this event was slightly older (i.e. Silurian or Early Devonian) in the west of the state cannot be precluded from the available evidence. The mineralizing event appears linked to the same thermal event as diorite (‘lamprophyre’) dykes and graite intrusions, acidic and basaltic volcanism, deformation and regional metamorphism. The limited amount of fluid chemistry data from Victorian gold deposits suggests low-salinity fluids which contained CO2, and variable methane, with homogenization temperatures from 150 to 300°C. The sulphur isotopic compositions of pyrite are mostly +1 to +5‰, with some values slightly above + 10‰ Mineralizing temperatures are inferred to be near 300°C, and pressure around 1–2 kbar. Sulphide-rich gold deposits in the east of Victoria, many of which contain significant base metals, contrast with the rest of the province (which contains mostly gold deposits with lower sulphide content), and this eastern area should possibly be viewed as separate from the Victorian gold province, per se. The benefits to Victorian gold mining coming from the 1980's global gold boom have been insignificant compared to the historical importance of the Victorian gold province. There seems little justification for assuming that all the large deposits have been discovered, and there is reason to believe new deposit styles may emerge to complement the limited current exploration activity which is focussed on the search for extensions and repeats of known deposit types. A major advance in scientific understanding of these oft neglected deposits could kindle a major exploration and production revival.

Kalgoorlie (Golden Mile plus Mt Charlotte) accounts for more than half the gold won from the 2000 deposits that comprise the widespread gold mineralization of the Archaean Yilgarn Block of Western Australia. It is also one of the largest gold deposits globally (> 1800 t Au production and reserves). An analysis of the various stages of formation of the Kalgoorlie deposits from source, through solution of gold in hydrothermal fluids and their infiltration along channelways, through to gold precipitation at depositional sites reveals several features that make Kalgoorlie favourably disposed as a site for a major gold deposit. However, none alone account for its anomalously large size. It is concluded that the giant Kalgoorlie gold deposit owes its size and unique position in Western Australian gold production history to favourable host-rock chemical composition, favourable host-rock thickness and mechanical properties, and to the favourable geometry of its host units and hosting greenstone belt, at favourable PT conditions within the greenschist facies. Collectively, these factors, together with proximity to the regional-scale Boulder-Lefroy fault, allowed highly focussed fluid flow on both a deposit and district scale. In this regard, it is significant that the structural setting of gold mineralization and the greenstone architecture at the largest Canadian Archaean deposit (Timmins) are remarkably similar to Kalgoorlie. Orebody size-distribution histograms for both the Yilgarn Block and the Abitibi Belt, Canada, are highly skewed, with extremely few giant deposits, but numerous deposits of ≤ 10 tonnes gold existing. This topology can be accounted for in terms of system efficiency. The probability of all six factors in the ore-forming system being optimized (high heat flow, large fluid reservoir, proximity to major structure, efficient fluid focussing, and large and compositionally favourable host lithology) is extremely small. In contrast, the probability of two processes coupling to give low efficiency is very large, resulting in numerous small deposits. Kalgoorlie appears to have formed by the same basic processes that formed the plethora of smaller Yilgarn gold deposits. However, it represents a mineralized domain in which there was a conjunction of a number of favourable parameters, both on a regional and deposit scale, that produced high system efficiency and allowed the effective fluid focussing and gold deposition that formed this giant deposit.

Sudoite, with almost ideal composition of Mg2Al4Si3O10(OH)8, has been found in the Khaki shale of the Witwatersrand Basin, South Africa. The sudoite occurs in aggregates of elongate bundles alternating with muscovite and rarely pyrophyllite, in an assemblage of sudoite-pyrophyllite-muscovite-quartz-rutile-pyrite. The stability of sudoite is attributed to a particularly Mg rich bulk rock composition, elevated sulphur activity, and metamorphic conditions just above the kaolinite to pyrophyllite transition. Nearby low variance chlorite, chloritoid-and pyrophyllite-bearing assemblages represent differing bulk rock compositions at potentially very similar metamorphic conditions. As the first find of sudoite in the Archaean, this occurrence in the Witwatersrand Basin extends the metamorphic setting and age in which sudoite is found to well beyond Phanerozoic belts of crustal thickening.

Low-salinity fluids (T > 200°C reduced S, modest CO2) and high geothermal gradients are common to many gold deposits and provinces. In contrast, host rocks, hosting structures, depth of formation (in the crust during deposition), subsequent metamorphic overprint, alteration mineralogy and isotopic signatures can vary dramatically within single deposits or provinces. Gold deposits with co-product base metals are an exception to the above comments, and probably relate to saline fluids.The low salinity fluids inferred for major gold-only deposits are not easily explained by seawater, basinal brines, meteoric fluid or common magmatic processes. In contrast, metamorphic devolatilisation of mafic/greywacke rocks is one effective way to produce low-salinity metamorphic fluids with characteristics matching the gold fluids. Such an origin also explains the link to geothermal gradients.The transition from chlorite—albite—carbonate assemblages to amphibole-plagioclase assemblages (commonly greenschist—amphibolite facies boundary) involves considerable loss of metamorphic fluid whose composition is buffered by the mineral assemblage, and is a function of P and T. This low salinity, H2O-CO2 fluid is evolved at T > 400°C commonly carries reduced sulphur, and may contain Au complexed with this sulphur. This auriferous fluid is likely to mix with other fluid types during times of elevated temperature, especially magmatic fluids at depth, and upper crustal fluids at higher levels.Gold deposits in Archaean greenstone belts exhibit good evidence of low salinity, H2O-CO2 fluids of T > 300°C these include examples from Canada, Australia, Brazil, Zimbabwe, India, and South Africa. Turbidite-hosted (slate-belt) deposits exhibit similar evidence for such fluids but commonly with appreciable CH4; the Victoria and Juneau (Alaska) goldfields are examples. The Witwatersrand goldfields also show evidence of low salinity, H2O-CO2 fluids carrying reduced sulphur and gold, but their distribution and timing are not well established. Epithermal (sensu lato) gold deposits have evidence for low salinity fluids carrying Au and S, but are much more diverse in character than those from the previously mentioned gold provinces: this probably arises from mixing of several fluid types at high crustal levels. Together these four types of gold provinces account for over 80% of the primary gold mined to date.

There is extensive textural evidence for sulphidation in the Middelvlei Reef of the Carletonville Goldfield, South Africa. Secondary iron sulphides have replaced sand- and pebble-sized elasts in conglomerate, as well as matrix material in both conglomerate and quartzite. Within the Middelvlei Reef there is a close correspondence between the intensity of sulphidation and areas of greater modal muscovite, and there is less sulphidation where chloritoid dominates. Both pyrite and minor pyrrhotite are widespread. The timing of the sulphidation appears to be post-depositional but the relative importance of diagenetic and syn-metamorphic processes cannot be determined using textural criteria. The absence of retrograde effects suggests that sulphidation occurred prior to or during peak metamorphism. The distribution and intensity of sulphidation and phyllosilicate alteration indicate that these processes may be more significant in the Witwatersrand than has generally been thought. These processes should be considered in all future genetic models.

Low-salinity, high-temperature (>200 degrees C), reducing fluids are poorly represented in ore-forming processes, yet appear to account for 90 percent of the primary gold mined to date. Such fluids have been implicated in the formation of many Archcan greenstone, Witwatersrand, slate belt, and epithermal gold deposits. These low-salinity auriferous fluids occur in variable time, space, lithological sequence, and tectonic settings and are most easily explained by metamorphic devolatilization of mafic and/or graywacke successions.Metamorphism of chlorite-calcite-albite-quartz assemblages (e.g., mafic or graywacke) at the greenschist-amphibolite facies boundary occurs around 480 degrees + or - 20 degrees C (for 3-5 kbars) and produces large volumes of low-salinity, H 2 O-CO 2 fluid similar in composition to those recorded in many gold deposits. Widespread pyrite in the above assemblage typically leads to elevated levels of dissolved sulfur. Such conditions are ideal for gold transport as a molecular Au-S complex and can lead to deposits rich in gold relative to silver and base metals.Deposition of gold occurs in the temperature range of 250 degrees to 400 degrees C and can be facilitated by lower temperature, lower sulfur activity, and changes in oxygen activity. Lower sulfur activity is most readily achieved by fluid interaction with Fe-rich host rocks with the formation of pyrite or pyrrhotite, whereas lower oxygen activity is readily achieved by interaction with carbonaceous metasediments.Low-salinity fluids produced by metamorphic devolatilization provide a common theme embracing the formation of many Archcan greenstone, slate belt, Witwatersrand, and potentially epithermal gold deposits. It involves the common features of high geothermal gradient, low-salinity fluid, high Au/Ag, high Au/base metals, and broad synchroneity with metamorphism. It also accounts for the considerable variability in hosting structures, detailed timing relations, host rocks, alteration assemblages, and geologic age of different deposits. Interestingly, the similarities are explained by deeper processes, the differences by processes at or near the depositional sites.

The giant Kalgoorlie gold deposit (1,200 metric tons Au recovered) is surrounded by a wide wall-rock alteration halo (>1 X 3 km). This alteration primarily represents the selectively pervasive formation of carbonate, chlorite, and localized intense pyrite and muscovite alteration, which all formed synchronously with gold mineralization. The carbonate alteration halo is interpreted as epigenetic because of evidence that carbonate-chlorite assemblages overprint regional metamorphic fabrics. Furthermore, the geometry of the alteration zone boundaries is controlled by ductile shear zones formed during regional deformation and the geochemistry of all the alteration zones suggests that one fluid was responsible for all assemblages. There is no evidence of intense, premetamorphic carbonation in the Kalgoorlie area.Kalgoorlie is situated in the highly auriferous Eastern Goldfields Province of the Yilgarn Block, within a narrow corridor (rift zone) that contains several large gold deposits. Regional metamorphism has produced upper greenschist facies assemblages but has retained original rock textures. At Kalgoorlie, the most important gold host is the Golden Mile Dolerite (80% of gold produced), a differentiated tholeiitic sill intruded between a lower mafic-ultramafic pile, including the Paringa Basalt (the next most important gold host), and overlying felsic rocks and metasediments.The Kalgoorlie syncline is the dominant tectonic structure in the main gold-producing area of the Golden Mile. Adjacent to this infold of metasediments, the Golden Mile Dolerite and Paringa Basalt are the site of a number of ductile shear zones which formed synchronously with the syncline. These zones, and the many smaller shear zones between them, have provided fluid channelways throughout the Golden Mile and are the site of most economic gold mineralization.Three alteration zones have been mapped in the Golden Mile Dolerite on the basis of the mineralogic siting of Fe (i.e., the chlorite, carbonate, and pyrite zones). The additional actinolite zone, representing the unaltered regional metamorphic mafic assemblage, has been converted to chlorite zone assemblages throughout the mined areas of the Golden Mile. The carbonate zone (bleached dolerite) can be traced for 2 km through the main mineralized parts of the Golden Mile and to the full depth of Lake View mine. The boundary between the chlorite and carbonate alteration zones is locally sharp and controlled by ductile shear zones but is generally strata-bound and preferentially developed in the more Fe-rich parts of the dolerite. The pyrite zone, which includes all economically gold-mineralized areas, is the site of the strongest muscovite alteration and is most intense and continuous in ductile shear zones.The chlorite zone (chlorite + calcite + or - ankerite) forms from mild carbonation of actinolite zone assemblages; the carbonate zone (ankerite + siderite) is a more local expression of this carbonation. Even after extensive carbonate alteration, relative major element abundances are little changed and the main added components are limited to CO 2 , S, K, and Rb. The pyrite zone (i.e., the gold lodes) has ankerite, muscovite, quartz, albite, pyrite, and minor siderite-bearing assemblages; it represents significant addition of sulfur. The three alteration zones not only reflect differing degrees of alteration intensity but are also strongly influenced by host-rock Fe/Mg and this ratio largely determines the stability of chlorite (favored by low Fe/Mg) and thus the distribution of the carbonate zone (favored by high Fe/Mg). Progressive deformation within the shear zones reworked existing mineralization.Kalgoorlie fits a metamorphic replacement model of gold formation by auriferous metamorphic fluids interacting with Fe-rich host rocks. Its great size is a function of a very well-developed fluid channelway system, via ductile shear zones and later fault zones, and the favorable chemical nature of the host rocks, especially the Golden Mile Dolerite. The full width of carbonation (chlorite and carbonate zones) reflects overlapping alteration envelopes around the closely spaced fluid pathways. The processes responsible for the formation of most of the gold at Kalgoorlie were essentially similar to those at several other Archean gold deposits, but on a larger scale.

Gold mineralization at the Hunt mine, Kambalda, is situated in a 20-m-thick, steeply dipping shear zone within a sequence of rather uniform metabasalts. Zonal alteration of the host rocks occurred around fractures which acted as conduits during gold mineralization. Limited major element mobility ensured assemblages of low variance over most of the alteration zone, providing a suitable environment for the application of thermodynamics in conjunction with detailed petrographic, geochemical (electron microprobe analysis of mineral compositions), and fluid inclusion studies (microthermometry).Equilibrium equations involving fluid components and alteration minerals are used to determine conditions during the alteration event, and in particular, activity gradients of components. Profiles across the schist zone show increasing f (sub H 2 S) around auriferous quartz veins (log f (sub H 2 S) = -0.6) but constant f (sub O 2 ) (log f (sub O 2 ) = -29.7). These results are near the magnetite-pyrrhotite-pyrite triple point for the inferred temperature of 350 degrees C and explain the pyrite-pyrrhotite-magnetite zonation sequence outward from veins. The profile of X (sub (CO 2 )) is very close to the value of 0.25 determined independently from fluid inclusions. The K (super +) /H (super +) profile shows increases adjacent to auriferous veins corresponding to the zone of biotite stability. Using a value for fluid salinity of 2 equiv wt percent NaCl, pH was calculated to be around 6.9.The solubility of gold in solution was estimated for the calculated conditions: for the complex Au(HS) (super -) 2 the maximum solubility is around 0.5 mg kg (super -1) (i.e., 0.5 ppm), for Au(Cl) (super -) 2 it is close to 10 (super -6) mg kg (super -1) , indicating the efficacy of sulfide transport of gold, compared to chloride transport, under these conditions. The calculated decrease in f (sub H 2 S) away from the auriferous quartz veins results in an 80 percent (or greater) decrease in gold solubility, suggesting hat the sulfidation of Fe-rich wall rocks is a geologically reasonable and very efficient mechanism for precipitating gold from solution at elevated temperatures.

Phase compositions in pelitic and mafic gneisses place tight constraints on pressure (ranging from 3 up to 6 kb), and, to a lesser extent, on temperature (500° up to 800° C) during prograde regional metamorphism of the Willyama Complex, Broken Hill, SE Australia. These limits allow an evaluation of water activity across the terrain using various equilibria in pelitic and mafic gneisses. The stability of cummingtonite and biotite over much of the terrain places upper limits on temperature, and the presence of syn-metamorphic partial melts in the metasediments places lower limits on a(H2O). Garnet-biotitesillimanite-K feldspar-quartz relations combined with the partial melting data suggest a decrease in water activity from near 1.0 in the lower grade zones to 0.5±0.2 in the Broken Hill — Little Broken Hill part of the two pyroxene zone. This result is compatible with less precise hornblende-orthopyroxene-clinopyroxene-quartz relations.

Gold deposits occur in greenstone belts world wide, and contribute to anomalously high gold production from Archaean terranes. As in other cratons, Archaean gold mineralization of Western Australia represents a complex array of deposit styles. Despite this, most deposits are clearly epigenetic, and large deposits have a number of features in common, including their strong structural controls, distinctive wallrock alteration (Fe-sulphide, K-mica±albite, CaMgFe carbonates), consistent metal associations (AuAgAsSbWB; low base metals), commonly Fe-rich host rocks, great depth extension and lack of appreciable vertical zonation. These shared characteristics, combined with their ubiquitous occurrence, indicate that Archaean gold deposits had a common origin related to the tectonic evolution of greenstone belts. Auriferous hydrothermal systems were broadly synchronous with regional metamorphism and emplacement of synkinematic granitoids and felsic (porphyry) intrusions. Although these gold systems involved low-salinity, lowdensity, reduced, near-neutral H 2 OCO 2 fluids carrying gold as reduced sulphur complexes, the origin of the fluids is equivocal. Most timing evidence and stable isotope data cannot distinguish metamorphic from magmatic (granitoid or felsic porphyry) orggins, but the lack of consistent spatial relationships between specific, volumetrically significant intrusive phases and large gold deposits in a number of cratons strongly favours metamorphic derivation of fluids. The metamorphic-replacement model for gold mineralization involves devolatilization of the lower portions of the greenstone pile, with high geothermal gradients inhibiting significant melting. CO 2 possibly formed by the decarbonation of early alteration, related to mantle degassing along crustal-scale, synbasinal fault zones. Auriferous fluids were channelled along greenstone-scale faults, in part developed during reactivation of crustal-scale faults in a strike-slip regime. Gold deposition occurred largely under greenschist facies conditions (about 300–400°C, 1–2 kb) in response to decreasing gold solubility with declining temperature. However, a major control on gold deposition was fluid/wallrock interaction. Many large deposits formed by sulphidation of Fe-rich host rocks, with synchronous deposition of Fe-sulphides and gold. However, the variable nature of gold-depositing reactions, including lowering of fO 2 and pH, allowed a multitude of small, and some large, deposits to form wherever that fluid circulation occurred. In consequence, several of the relatively small deposits currently worked from open pit are hosted by ultramafic and felsic rocks. There are few constraints on the source of components (Au, S, K, CO 2 ) added to gold deposits, but even giant deposits such as the Golden Mile, Kalgoorlie could have formed from a realistic greenstone source volume (ca. 8×8×5 km). Convective circulation of fluids could have contributed to the generation of high fluid-rock ratios. On the regional scale, the markedly heterogeneous distribution of large gold deposits, gold productivity and host rocks to deposits can be accommodated by the metamorphic-replacement model. The most favourable conditions for development of auriferous hydrothermal systems operated in younger (ca. 2.7±0.1 Ga) rift-phase greenstones where greatest extension and crustal thinning produced high geothermal gradients, crustal-scale synbasinal faults, and rapid extrusion and burial of volcanics, including abundant komatiites. Iron-rich tholeiitic basalts and dolerites were preferred host rocks for large gold deposits. The least favourable conditions existed in older (ca. 3.5-3.4 Ga) platformphase greenstones, where gentle sagging on submerged continental crust produced eruption of mainly mafic volcanics with few komatiites, commonly in very shallow-water environments. This allowed intense synvolcanic alteration of both gold source rocks and potential host rocks. The generally smaller gold deposits formed mainly in ultramafic or greywacke hosts. Younger (ca. 3.0 Ga) platform-phase greenstones appear intermediate in nature but, unlike other greenstones, have significant epigenetic gold deposits in originally oxide-facies BIF, which were deposited on relatively deep-water platforms. Similar controls appear to exist on a world scale, with gold mineralization peaking at ca. 2.7±0.1 Ga in response to development of major rift zones in thickened, relatively mature continental crust. Interestingly, the giant Witwatersrand goldfield formed at about the same time.

Mineralogical and geochemical evidence indicates that partial melting and desulphidation have occurred in the Big Bell gold deposit. Decarbonation may also have occurred, to account for the lack of a carbonate alteration halo; this is compatible with the present data, but difficult to test. The Big Bell deposit consists of auriferous sulphide-bearing (‘lode’;) schists with muscovite and K-feldspar, and surrounding biotite schists, all derived by intense premetamorphic alteration of rocks of mafic composition. Assemblages which include cordierite-sillimanite-K-feldspar-garnet-biotite-quartz suggest peak metamorphic conditions of 4–5 kbar, and 650–700° C, based on phase relations, geobaro-meters and garnet-biotite Fe-Mg exchange partitioning. Partial melting occurred at peak metamorphism, particularly in the altered mafic rocks in and around the deposit, and its occurrence may have been essential to the preservation of the deposit. Melting greatly limited the importance of devolatilization reactions, resulting in negligible aqueous fluids and no means of removing appreciable gold. Minor gold loss may have accompanied desulphidation. A diversity of complex metamorphic assemblages occurs around the mine, compared to the assemblages developed regionally; variable bulk rock composition influences this contrast, but there is no evidence of higher metamorphic grades at the mine, nor that this might have been the prime control on the different assemblages in this narrow belt. It is suggested that the Big Bell and Hemlo deposits are the higher metamorphic grade equivalents of the more abundant greenschist facies gold deposits within Archaean greenstone belts. This interpretation is favoured by the host rock setting and geochemical characteristics of Big Bell. Alternative models that suggest that this class of deposit is a new type must account for the absence of high-grade equivalents of the greenschist facies deposits and also the lack of low-grade equivalents of the Big Bell/Hemlo type. Archaean gold deposits in high-grade metamorphic terrains have undergone a series of processes that are not recorded in the more typical gold deposits of the greenschist facies.

Mesoscopic and microscopic analysis of the structurally controlled gold mineralization of the Golden Mile, Kalgoorlie, has demonstrated that the mineralogically simple gold pyrite mineralization style (70% of production) is restricted to an anastomosing system of narrow high strain zones (ductile shear zones) and to related tension fractures in the intervening unstrained rock. Virtually all ductile shear zone lode structures occur in the Golden Mile Dolerite because it was essentially mechanically isotropic prior to deformation. It is proposed that during the near-upright regional-folding event numerous ductile shear zones were established in an approximately axial planar relationship to the folds, with synchronous conjugate shear zones developed at low angles to this direction. A dynamic analysis shows that the tensional cross lodes can be related geometrically to the three ductile shear zones. Golden Mile-style mineralization is syntectonic and was not deposited posttectonically in structurally prepared sites. Microtextural analysis demonstrates that minerals associated with gold emplacement are strongly modified during strain within the ductile shear zones by ductile deformation dominated by mass transfer processes. In particular there is abundant evidence of crack-seal-type microfracturing within and marginal to grains and in microscopic transgranular sites; this deformation style contributed prominently to the bulk plastic flow of the shear zones. It is proposed that high fluid pressure gradients were established at each crack-seal cycle leading to much enhanced permeability and to considerable focusing of fluid flow through the narrow foliated zones. This structurally induced infiltration mechanism allowed the ore fluid to pervade the deforming zones and also controlled the straining process by moving dissolved mineral components from highly stressed contacts to sites of deposition. Thus the mineralization was introduced during the ductile phase of shear zone evolution, and though mesoscopic veins are present in the ductile shear zones, they are not a necessary part of the gold emplacement event: the majority of veins formed late in the tectonic history. Gold emplacement in the ductile shear zones was controlled by microscopic deformation processes and ductile shear zones are good exploration targets even in the absence of macro- to mesoscopic dilation sites.

Genetic models should not only attempt to explain features of Archean gold mineralization on the scale of mines or districts but should also explain its marked spatial and temporal heterogeneity in greenstone belts. The application of genetic models developed with such a philosophy can have predictive capacity in gold exploration.The metamorphic model, one of a number of current genetic models for Archean gold deposits, involves generation of auriferous H 2 O-CO 2 fluids by devolatilization of the stratigraphically lowermost greenstones during metamorphism and broadly synchronous synkinematic granitoid intrusion. Crustal-scale fault-shear systems are required to induce channeled fluid flow and to focus ore fluids toward suitable host rocks and structural sites. Declining temperature below the amphibolite-greenschist transition, combined with sulfidation of Fe-rich host rocks, is a particularly efficient Au-depositing mechanism. The model suggests that the major controlling factors (in declining order of importance) are: (1) high temperature-low pressure, generally low-strain, metamorphic regimes, (2) crustal-scale fault systems, (3) unaltered Fe-rich rocks (tholeiites, banded iron-formation) at a suitable stratigraphic-metamorphic level, and (4) abundant Au-enriched komatiites and S-rich sediments in source regions.The low-pressure metamorphic conditions combined with the occurrence of elongate high strain zones in most, if not all, generally low strain greenstone belts may explain the ubiquitous occurrence of gold deposits in Archean terranes. All four favorable parameters appear to act in concert to produce the largest gold deposits and most abundant gold mineralization in 2.7 + or - 0.1-b.y.-old rift-phase greenstone belts, interpreted to have formed under conditions of high crustal extension and extensive crustal thinning (i.e., with predicted high geothermal gradients, high fault density, limited synvolcanic alteration, and abundant komatiites and sulfidic sediment horizons). More widespread, commonly older, platform-phase greenstone belts, which formed under conditions of lesser crustal extension and thinning, are less mineralized, as predicted by the metamorphic model.The ability of the model to explain the regional distribution of known Archean gold deposits lends credence to its predictive capacity. For example: (1) greenstone belts in any one craton can be broadly ranked in order of likely relative gold potential, (2) the likelihood of discovery of large deposits in poorly exposed or poorly explored belts of known type can be estimated, (3) areas with the highest potential can be defined in rift-phase greenstone belts by the coincidence of critical parameters such as specific host-rock sequences and metamorphic facies, (4) prospective structural sites for gold mineralization can be postulated in platform-phase greenstone belts affected by diapiric tectonics, and (5) the possible spacing of large deposits can be estimated. Further, the potential use of the model in exploration emphasizes serious gaps in our knowledge of parameters crucial to understanding greenstone belt metallogeny, including alteration patterns, metamorphic P-T paths, and regional-scale structural regimes.