The Porgera gold deposit, Papua New Guinea : geology, geochemistry and geochronology

Abstract

The Porgera gold deposit, located in the highlands of Papua New Guinea (PNG), contains ~410 tonnes Au and ~890 tonnes Ag, distributed between a large lower-grade ore zone (78.6 million tonnes, 3.5 g/tonne Au, 9.9 g/tonne Ag), and a smaller highgrade zone (5.0 million tonnes, 26.5 g/tonne Au, 22.1 g/tonne Ag). Lower-grade mineralization occurs as stockworks and disseminations of auriferous arsenical pyrite associated with strong sericite-carbonate alteration, which cross-cut and overprint a suite of epizonal mafic stocks and dykes of the Porgera Intrusive Complex (PIC) and their sedimentary host rocks (Jurassic-Cretaceous shelf sediments). Minor free gold also occurs in base metal-sulphide veins associated with these alteration zones. Later, high-grade mineralization (locally up to 3000 g/tonne Au) occurs in quartz-oscoelite veins associated with the Roamane Fault (an extensional structure which cross-cuts the intrusive complex and the earlier dissemiriated ore). Abundant visible gold occurs with Au-Ag-Ag-tellurides in these epithermal-style, vuggy, banded veins. Deposition of both types of ore is shown by K-Ar dating of sericite (illite) and roscoelite to have occurred within 1 Ma of the time of emplacement of the PIC at 6.0 ± 0.3 Ma (2a; K-Ar dating of igneous biotite, and 40Ar/39 Ar dating of hornblende). Geochemical, isotopic, and petrographic studies of the PIC indicate that the intrusions represent a comagmatic, volatile-rich alkali basalt/gabbro - hawaiite (trachybasalt) -mugearite (basaltic trachyandesite) fractionation suite, derived from a larger fractionating magma chamber located deeper in the upper crust (aeromagnetic evidence). The intrusions are medium- to coarse-grained, and textures vary from porphyritic to ophitic. Mafic rocks contain olivine (pseudomorphs) and Cr-rich diopside phenocrysts, whereas hornblende (titanian magnesio-hastingsite), Ti-rich salite, and plagioclase phenocrysts occur in hawaiites and mugearites. Fluor-apatite and magnetite (Cr-rich in mafic rocks) occur as phenocrysts throughout the suite. High Fe3+/Fe2+ ratios in whole-rock samples, pyroxenes and amphiboles, and the presence of primary chromite/magnetite microphenocrysts indicate that the magma crystallized under conditions of high fo2. Least-evolved samples are characterized by low Ba/La (8 to 10), La/Nb (0.6 to 0.7) and Sr/Nd (~25) ratios, and LREE-enrichments ([La/Yb]cn = 15 to 19), which are similar to those of intra-plate alkali basalts, and distinguish the PIC from other Late Tertiary K-rich alkaline and calc-alkaline volcanics and intrusions found on the PNG mainland. Isotopic compositions are relatively depleted (ENd= +6, 87Sr/86Sr = 0.7035, 206Pb/204Pb = 18.64, 207Pb/204Pb = 15.55, 208Pb/204Pb = 38.45), and evidence is found for only limited amounts of crustal contamination. These data suggest derivation of the Porgera magmas by partial melting of a garnet-lherzolite source in the upper mantle. The incompatible element-, volatile-rich nature of the magma suggests that the mantle source region had undergone metasomatic-enrichment prior to melting ( < 0.5 Ga), and the timing of magmatism suggests that both metasomatism and melting may have been related to the elimination of an oceanic microplate segment by double subduction beneath the Australasian (PNG) and Pacific (Bismarck Sea) plates. A model is proposed which involves metasomatism in the back-arc asthenosphere by fluids or melts derived from the subducted slab at depths below ~150 km. Isotopic tracing studies in the ore deposit indicate that hydrothermal Pb and Sr were derived from a mixture of igneous and sedimentary sources, located within the Jurassic Om Formation (carbonaceous, pyritic siltstones). These fluids carried K, Rb, Mn, S, C02 and other components including Au and Ag into depositional zones in the overlying Cretaceous Chim Formation. Mass balance calculations indicate a significant magmatic involvement in the source of hydrothermal Pb, but show that Sr was largely derived by leaching of the sedimentary sequence. Analyses of precious metal abundances in unaltered intrusive rocks-and sediments show that neither rock-type represents a significantly gold-enriched protore, but evidence for the evolution of a volatile phase during crystallization of the magma suggests that Au and other elements may have been partitioned into a magmatic fluid. It is suggested that mixing between this fluid and warm, reduced, sulphide-rich groundwaters circulating in the Om Formation sediments resulted in rapid deposition of base metals, but gold was retained in solution as a bisulphide complex until precipitation at higher levels in the Chim Formation. Gold deposition was controlled by a combination of cooling, wallrock reaction, fluid mixing, and/or boiling, which resulted in destabilization of bisulphide complexes. A late influx of fresh magma into the parental magma chamber, probably triggered by tectonic activity, resulted in the emplacement of a suite of feldspar porphyry dykes, and the release of a final pulse of hydrothermal fluid. These fluids ascended along late faults and subsidiary structures, and rich deposits of gold were formed where they boiled or mixed with cool, descending groundwaters.