Finite Element Heat and Fluid-Flow Computer Simulations of a Deep Ultramafic Sill Model for the Giant Kidd Creek Volcanic-Associated Massive Sulfide Deposit, Abitibi Subprovince, Canada

Abstract The giant Kidd Creek volcanic-associated massive sulfide deposit in the western Abitibi subprovince of Canada is unique in that it is a single deposit rather than one of many in a district, and has a footwall comprising proximal, high-temperature, high silica rhyolites intercalated with komatiite flows. Classic volcanic associated massive sulfide models with a relatively shallow, synvolcanic sill cannot account for its size or singularity. Finite-element heat and fluid-flow computer simulations of a two-dimensional physical model are presented here that can explain how such a deposit could be produced. The physical model has (1) a relatively deep ultramafic sill as the principal heat source, (2) a rhyolite magma conduit that is coincident with a relatively permeable fault, and (3) a less permeable basaltic substrate. All of the computer simulations employ a temperature-dependent, permeability enhancing and reducing thermal cracking front and have a 1.7-km-thick ultramafic sill with its base at 15 km which is set at 1,650°C for 50,000 yr and then allowed to cool. In the preferred computer simulation, rhyolite magma is injected in the magma conduit episodically during the first 10,000 yr. With these parameters, a single hydrothermal convection cell is established that vents in one location at>200°C for ∼650,000 yr, with an average venting temperature of ∼300°C and an average venting rate of ∼175 cc/cm2/yr. Venting temperatures and rates drop rapidly thereafter. Approximately 15 million metric tons (Mt) of Cu and Zn in the deposit are provided considering a hy-drothermal fluid that precipitates 100 ppm Cu + Zn, a 0.6-km2 vent area within a catchment graben, and a 22 percent depositional efficiency. The total energy in the hydrothermal fluid expelled is ∼1.53 X 1020 cal, or 10.4 MW/km of graben axis. This estimate of hydrothermal energy output per kilometer of rift axis is one-half to two-thirds that of the Taupo volcanic rift zone of New Zealand and about twice that of the average high-temperature hydrothermal flux at the midocean ridge axes. In the simulations, a thermal cracking front initially forms at the top of the sill at a ∼13-km depth, then migrates to a ∼8.5-km depth as conductive heat from the sill penetrates into the overlying substrate, and finally collapses down to 13 km as the heat from the sill is mined out. Thus, hydrothermal fluids doubly scour metals from a deep basaltic source region. Thermal cracking also occurs adjacent to and within the rhyolite conduit during and after magma injections, which helps establish hydrothermal venting in one area. Simulations without the rhyolite injections indicate that stable hydrothermal venting at>200°C is delayed by ∼100,000 yr and produces ∼10 percent less hydrothermal fluid and energy than with the rhyolite injections. Other simulations indicate that higher permeability substrates lead to higher venting rates but with lower venting temperatures and shorter durations. Lower permeability substrates would prevent hydrothermal circulation. This suggests that there is a permeability window for optimal, protracted, high-temperature venting and optimum scavenging of metals from the basaltic substrate. These computer simulations provide evidence that singular, large volcanic associated massive sulfide deposits of Kidd Creek dimensions can occur where deep ultramafic sills provide heat. The presence of proximal high-temperature, high silica rhyolites which form by partial melting of tholeiitic basalt above the garnet stability field indicates anomalously hot crust possibly related to ultramafic sills and can help delineate vent source areas. Deep, crustal-scale fractures in relatively primitive, ultramafic, or picrite-bear-ing terrane such as primitive greenstone belts, oceanic plateaus, or thickened oceanic rifts are also favorable locations for voluminous hydrothermal venting and large volcanic associated massive sulfide deposits; such fractures are expected during incipient rifting or mantle plume upwelling in oceanic terrane, away from active midocean rifts.