Nicholas H.S. Oliver

Synorogenic hydrothermal origin for giant Hamersley iron oxide ore bodies

Geologic mapping, basin analysis, and calculated fluid compositions indicate that giant orebodies of microplaty hematite, and possibly martite-goethite, in the Hamersley province of Western Australia, were formed by heated fluids driven by early Paleoproterozoic orogenesis. Detrital grains of microplaty hematite in the McGrath trough, a foreland basin in front of the northward-advancing Ophthalmian fold belt constrain the age of the earliest microplaty hematite ore formation to 200 °C and locally up to 400 °C were involved. Regional circulation of hydrothermal fluids, including heated surface water, through reduced banded iron formations occurred during or soon after the Ophthalmian orogeny. We speculate that martite-goethite orebodies, previously considered Mesozoic–Cenozoic, could also be related to heated Paleoproterozoic meteoric fluids migrating northward away from the Ophthalmian fold belt.

Numerical models of extensional deformation, heat transfer, and fluid flows across basement-cover interfaces during basin-related mineralization

Fluid circulation within low-permeability basement rocks has been proposed to occur beneath many sediment-hosted mineral deposits, in some cases contributing substantial metals or sulfur to the deposits in overlying cover sequences. However, mechanisms proposed for fluid transport and mass transfer within and through basement rocks are diverse, some models appealing to thermal circulation but others appealing more to deformation- or topography-driven flow. We address some of these issues here by a series of numerical models designed to compare and then couple thermally and mechanically driven fluid flow (and incorporate temperature-dependent fluid properties), starting with generic problems and then using a simulation of coupled deformation, heat transfer, and fluid flow that may be applicable to the formation of Mount Isa-style Pb-Zn ores and other extension-related basinal deposits. Results from deformation-only models show that downward penetration of near-surface fluids into relatively low permeability basement rocks may occur along fault zones at high strain rates during extension, because local deformation rates may exceed the capacity for fluid to move through the basement rocks due to their low permeability, leading to periods of underpressure. For our thermal fluid-flow models, in the absence of deformation and with elevated basal heat flows, large differences in basement and cover permeability tend to restrict thermal convection to the permeable units. Downflow into low-permeability basement may occur by a reduction of the permeability of cover sequences, because larger convection cells are possible as permeability approaches common, optimal values throughout the rock mass. The normal reduction in porosity and permeability of cover sequences with burial may thus lead to progressively deepening convection cells and an enhanced potential for extraction of components from basement rocks. Long-lived, stable convection is generated with ≤2 order of magnitude permeability difference between basement and cover. Such convection has the potential to lead to near-surface mineralization (e.g., sediment-hosted syngenetic or diagenetic deposits), particularly if an initial overpressure stimulates convection cells toward upflow along basin-bounding faults. These models also serve to indicate the inadequacy of models that do not incorporate thermal dependencies of fluid viscosity and density, because the upward fluid velocity generated by buoyancy is of the same order of magnitude as the downward fluid velocity generated by extension-related underpressure in models that do not incorporate these properties. In numerical models of coupled deformation, heat transfer and fluid flow in which high basal heat flow is coupled with extensional deformation, the effects of the deformation dominate flow regimes, rather than the thermal structure. A model with initial heating and fluid flow established large convection cells with basement fluid circulation, prior to deformation being incorporated. The convection cells are effectively destroyed by extension at geologically reasonable strain rates around 10–14s–1, with surface fluids driven downward and meeting remnants of the decaying convection deep in the system. This simulation provides a possible solution for mixing of near-surface and deep fluids in unconformity-related U deposits and Olympic Dam-style iron oxide Cu-Au deposits. Geological models for shale-hosted base metal deposits (e.g., Mount Isa Zn-Pb) appeal to transitions from active rifting to blanketing by mineralized sag-phase shales, requiring reduction or cessation of extension with time. We simulate this here by stopping the deformation component of the coupled model and allowing the heating and fluid-flow parts to continue. Initial or periodic fluid overpressures (140% of hydrostatic) applied at the base of our coupled numerical models during extension (rift phase) cause initial upflow along faults and sufficient heat advection to generate steep near surface thermal gradients. When deformation ceases, convection progressively deepens with time, but upflow continues along faults, producing perfect conditions for exhalation of fluids that have circulated through basement. From all of the coupled models, we infer that active extension or extensional reactivation of basin-bounding faults is generally destructive with respect to potential fluid upflow and generation of near-surface deposits. Exhalative or other near-surface ores are likely to form when extension ceases and the thermal structure becomes the driver of fluid flow.

From banded iron-formation to iron ore: geochemical and mineralogical constraints from across the Hamersley Province, Western Australia

Several major iron ore deposits occur in deformed regions of the Hamersley Province, Western Australia, where banded iron-formation (BIF) of the Dales Gorge Member has been converted to martite and microplaty hematite. The genesis of these high-grade hematite ores remains controversial, in part because no study has systematically documented variations on the chemistry and mineralogy of stratigraphically equivalent rocks from undeformed regions into the deposits. In this study, we examine the powder colour, chemistry and mineralogy of 177 samples of the Dales Gorge Member and surrounding shales from the type section near Wittenoom and the Mt. Whaleback mine near Newman. Profound chemical and mineralogical changes suggest that after early diagenesis, low-grade metamorphism converted clays in black shales to stilpnomelane and talc. Coincident with or following these changes, reduced metamorphic fluids altered phyllosilicates and K-feldspar in these rocks to clinochlore and muscovite around Mt. Whaleback. These metamorphic fluids did not significantly affect BIF. However, subsequent acidic and oxidizing fluids around Mt. Whaleback converted magnetite to martite and dissolved carbonates and silicates from BIF. In black shales, these fluids also dissolved quartz and converted clinochlore and muscovite to hematite and kaolinite, respectively. Late in the paragenetic sequence, BIF or altered BIF was converted to highly porous high-grade hematite ore by the dissolution of Si, a process not requiring iron addition. Our observations in and around Mt. Whaleback suggest that BIF was oxidized prior to silica removal so we infer that the bulk of this oxidation was related to ore formation. However, regionally, this sequence is probably more complicated because rocks previously described at Mt. Tom Price show a stage involving carbonate replacement of silica prior to oxidation that is not evident at Mt. Whaleback. Also, the oxidation stage could (at least in part) have formed by weathering sometime since the Proterozoic. In any case, (1) no single process can produce all of the altered rocks at Mt. Whaleback, (2) oxidation of magnetite to hematite can occur independently of silica removal or replacement and (3) the main mineralization event postdates metamorphism. A model explaining the differences between these two deposits requires either a local carbonate addition step at Mt. Tom Price, or the complete removal (by oxidation) of previous carbonate–magnetite bearing ore at Mt. Whaleback.

Ostwald ripening as a possible mechanism for zircon overgrowth formation during anatexis: theoretical constraints, a numerical model, and its application to pelitic migmatites of the Tickalara Metamorphics, northwestern Australia

A fundamental dichotomy exists between the low solubility of zircon in peraluminous melt predicted by experimental and geochemical studies and the large volume proportions of zircon overgrowths formed during high-temperature metamorphism and anatexis that are revealed by cathodoluminescence imaging. We investigate the potential of Ostwald ripening as a possible mechanism for overgrowth formation by presenting a numerical solution to the continuity equation governing open system, diffusion rate-limited Ostwald ripening in a zircon-saturated melt. Application of the model to a typical (log-normal) initial zircon crystal size distribution (CSD) suggests that despite uncertainties associated with the interfacial free energy of zircon, significant grain coarsening is possible via this mechanism under geological conditions and time scales relevant to high-grade metamorphism. Primary influences on the rate at which Ostwald ripening proceeds are (i) the temperature of the system, (ii) the duration of the time interval for which the system is above its solidus, and (iii) the nature of the initial (premelting) zircon CSD. To test the viability of the model, we examine zircon CSDs from three high-grade pelitic migmatites of the Tickalara Metamorphics (northwestern Australia), assuming that zircon crystals hosted by melanosome biotite were permanently occluded from the melt (and therefore approximate the premelting CSD). The model predicts that within 1 to 2 Ma, these biotite-hosted zircon CSDs will evolve into the observed leucosome-hosted zircon CSDs via melt-present Ostwald ripening, under geological conditions applicable to peak metamorphism. Although we have not conclusively demonstrated that Ostwald ripening contributed to changes in zircon CSDs during anatexis of the Tickalara metapelites, our results suggest that Ostwald ripening is a viable mechanism for zircon volume transfer in a zircon-saturated melt and capable of playing a significant role in overgrowth formation in rocks where the total volume of zircon overgrowths substantially exceeds the concentration of zircon dissolvable in the coexisting melt.

Insights into the genesis and diversity of epigenetic Cu – Au mineralisation in the Cloncurry district, Mt Isa Inlier, northwest Queensland

The Proterozoic rocks of the Cloncurry district preserve the effects of some of the worlds largest hydrothermal systems associated with extensive albitisation, brecciation and Na – Ca alteration. These hydrothermal systems are broadly coeval with magmatism, and also host numerous structurally controlled Fe oxide and Cu – Au deposits (ca 1.60 Ga, 1.55 – 1.50 Ga). Fluid-inclusion, stable-isotope, and geochemical data from Cu – Au deposits indicate that the ore-forming fluids were high-T (>300 – 500°C), highly saline (>26 – 70 wt % NaClequiv), typically CO2-bearing, and are mainly considered to be sourced by crystallising intrusions with contributions from other fluid sources and/or host rocks. Fe oxide and Cu – Au mineralisation in the district exhibit a range of interrelationships based upon the metal endowment, relative timing of Fe oxides and sulfides, and Cu:Au ratio. These interrelationships may be divided into four categories: (i) barren magnetite and/or hematite ironstones; (ii) Fe oxide-hosted Cu – Au mineralisation, where relatively Au-rich ore associated with pyrite and hematite overprints older magnetite-rich rocks; (iii) Fe oxide Cu – Au mineralisation, where both Fe oxides and Cu – Au mineralisation are cogenetically deposited; and (iv) Fe oxide-poor Cu – Au mineralisation, where relative Cu-rich mineralisation is associated with pyrrhotite and rare magnetite, and is hosted in relatively reduced rocks such as carbonaceous metasedimentary rocks. These categories reflect variations in fluid redox, f S, aFe, and temperature, as well as host-rock composition. The spectrum from Cu-rich to Au-rich mineralisation is a common phenomenon in Fe oxide – Cu – Au districts and predominantly reflects an increase in the redox of the ore-forming system. The apparent relationship between pH and metal solubility at different redox conditions suggests that Cu – Au mineralisation occurred as a result of decreasing fluid acidity by wall-rock reaction at the site of ore deposition, or potentially by mixing of fluids of different acidity. Fluid mixing provides an effective means to produce high-grade ore deposits via changing pH, cooling, and dilution in hydrothermal systems involving little wall-rock interaction.

Evidence for Reaction-induced Microfracturing in Granulite Facies Migmatites

Water-undersaturated partial melting at granulite facies conditions, followed by accumulation and upward migration of the resulting granitic melt, is one of the principle causes of crustal differentiation. Microfracturing caused by small, positive volume changes associated with water-undersaturated melting reactions may facilitate rapid extraction of melts from melting sites. We have successfully imaged annealed microfractures in residual quartz from three granulite facies migmatites. Annealed microfractures have been discovered in (1) quartz inclusions in peritectic garnet; (2) quartz inclusions in K-feldspar; and (3) quartz occurring as entrained grains. In the latter, oscillatory zoned quartz has overgrown fractures in the residual core, indicating that fracturing occurred prior to melt crystallization and that postanatectic volume changes are not responsible for the observed fracturing. Fractures are generally <5 µm wide, are parallel sided, and have low sinuosity. Crack densities are two to three times higher than experimental studies involving muscovite melting and approach theoretical interconnectivity thresholds for percolation, implying that reaction-induced microfracturing is a viable mechanism contributing to melt interconnectivity during anatexis under granulite facies conditions.

Fluid mixing versus unmixing as an ore-forming process in theCloncurry Fe-oxide-CuAu District, NW Queensland, Australia: evidence from fluid inclusions

Abstract Fluid mixing and/or unmixing (including boiling) are thought to be important mechanisms of mineralisation in copper-golddeposits. Detailed fluid-inclusion studies of regional sodic (-calcic) alteration and local mineralisation in the Cloncurry Fe-oxide-Cu Au District, NW Queensland, suggest that both fluid mixing and unmixing occurred in these giant mineralised hydrothermal systems. In some cases, the primary character of coexisting multisolid, hypersaline brine inclusions and CO 2 - or vapour-rich inclusions, the latter crosscut by late Ca- and Na-rich fluid inclusions, indicate that fluid mixing probably occurred subsequent to fluid unmixing and finally resulted in Cu Au mineralisation. However, the relationship between hypersaline brines and CO 2 , which was believed to result from an unmixing of a magma-derived H 2 O CO 2 NaCl ± CaCl 2 fluid (see [Miner. Depos. 36 (2001) 93] and references therein), is rather complex as some hypersaline brine inclusions obviously predate CO 2 inclusions.

Airborne hyperspectral imaging of hydrothermal alteration zones in granitoids of the Eastern Fold Belt, Mount Isa Inlier, Australia

ABSTRACT Hyperspectral remote sensing data from the Eastern Fold Belt, Mount Isa Inlier, Australia were compared with petrographic and geochemical studies to map the spatial extension and compositional variations of Proterozoic granitoids and endoskarns as well as hydrothermal alteration patterns in adjoining metasedimentary successions. Detailed spatial analysis of spectral remote sensing data shows an almost circular alteration zoning in the Mallee Gap Granite, which was emplaced during a late phase of the Mesoproterozoic Williams event. A combination of hyperspectral images, such as white mica, kaolin and MgOH products, were used to map the alteration zoning. The formation of the endoskarn is presumably related to autometasomatism and interaction with fluids released from the country rocks during a late phase of the emplacement. The intrusion of the Mallee Gap Granite has only a local control on the hydrothermal alteration, but high potassic granites of the southern Mount Angelay Granite might have expelled oxidized mineralizing fluids and possibly had a major impact on regional scale alteration. Hyperspectral remote sensing data may be used to estimate the imprint of single igneous bodies on the Mesoproterozoic hydrothermal evolution of the Eastern Fold Belt and are important for the study of ore-forming hydrothermal processes in general.

Basement metal scavenging during basin evolution:Cambrian and Proterozoic interaction at the Century ZnPbAg Deposit, Northern Australia

In some Phanerozoic base metal provinces floored by continental crust (Mississippi Valley, Irish midlands), controversy has arisen concerning whether basement rocks contribute significantly to the metal (and isotope) budgets, as opposed to surface or basinal fluids and rocks. For the Mesoproterozoic Century Zn-Pb-Ag deposit, we demonstrate that fluids associated with compaction of overlying Cambrian limestones penetrated the deposit, scavenged lead and zinc, and redeposited them in the limestones. This circulation was accompanied by penetration of carbonate breccia dykes into the basement, formation of shallow dipping shear zones and folds in the cover, and development of sulphide-bearing veins and stylolites. The process was facilitated by syn- or post-Cambrian reactivation of the Termite Range Fault, which was previously active during the Proterozoic ore deposition.

Early tectonic dewatering and brecciation on the overturned sequence at Marble Bar, Pilbara Craton, Western Australia: dome-related or not?

Cataclastic breccias and hydrothermal fault arrays of likely c. 3400 Ma timing are well developed and exceptionally well exposed in the Marble Bar Chert Member of the Pilbara Craton. Brecciation involved centimetre- to metre-scale clast transport distances, in breccia zones up to 5 m wide, cutting the c. 60 m thick chert in a series of right-lateral fault zones. Our observations of downward facing pillow basalts, the geometry of the breccias, and oxygen isotope data for rocks and the breccia matrix suggest the rocks were at least steeply overturned on this flank of the Mt Edgar Dome prior to brecciation. The breccias are inferred to represent steep conjugate fault zones developed by local transtension. The history of overturning and brecciation predates the formation of dome-related regional foliation and metamorphism, and therefore occurred between 3460 and 3320 Ma, the established ages for deposition of the underlying Duffer Formation and intrusion of the Mt Edgar Batholith respectively. Local overturning of the Marble Bar sequence prior to both brecciation. and the main phase of dome formation suggests a protracted deformation history for this segment of the Pilbara Craton. The transtensional movement along the breccias may be representative of strain accommodation accompanying an early doming phase, or could be a deformation event that developed independently of doming. Fluids involved in brecciation were most likely formation waters expelled from the cherts and basalts in response to overpressuring induced by the overturning and progressive burial.