Understanding controls on hydrothermal dolomitisation: insights from 3D reactive transport modelling of geothermal convection

Abstract

Abstract. The dominant paradigm for petrogenesis of high-temperature fault-controlled dolomite, widely known as “hydrothermal dolomite” (HTD), invokes upwelling of hot fluid along faulted and fractured conduits from a deep over-pressured aquifer. However, this model has several inherent ambiguities with respect to fluid sources and their dolomitisation potential, as well as mechanisms for delivering enough of these reactive fluids to form substantial volumes of dolomite. Here, we use generic 2D and 3D reactive transport simulations of a single transmissive fault system to evaluate an alternative conceptual model whereby dolomitisation is driven by seawater being drawn down into the subsurface and heated. We examine the evolution of fluid chemistry and the distribution of diagenetic alteration, including predictions of the rate, distribution, and temperature of HTD formation, and consider the possible contribution of this process to the Mg budget of the world's oceans. The simulations suggest that it is possible for convection of seawater along the fault damage zone to form massive dolomite bodies that extend hundreds of metres vertically and along the fault within a timescale of a few tens of thousands of years, with no significant alteration of the country rock. Dolomitisation occurs as a gradient reaction by replacement of host limestones and minor dolomite cementation, and it results in the discharge of Mg2+-poor, Ca2+-rich fluids to the sea floor. Fluids sourced from the basement contribute to the transport of heat that is key for overcoming kinetic limitations to dolomitisation, but the entrained seawater provides the Mg2+ to drive the reaction. Dolomite fronts are sharper on the “up-flow” margin where Mg2+-rich fluids first reach the threshold temperature for dolomitisation, and the “down-flow” dolomite front tends to be broader as the fluid is depleted in Mg2+ by prior dolomitisation. The model demonstrates spatial contrasts in the temperature of dolomitisation and the relative contribution of seawater and basement-derived fluids which are also commonly observed in natural fault-controlled dolomites. In the past, such variations have been interpreted in terms of major shifts in the system driving dolomitisation. Our simulations demonstrate that such changes may also be a product of emergent behaviour within a relatively stable system, with areas that are dolomitised more slowly recording the effect of changes in fluid flow, heat, and solute transport that occur in response to diagenetic permeability modification. Overall, our models robustly demonstrate that high-temperature fault-controlled dolomite bodies can form from mixed convection and act as a sink for Mg in the circulating seawaters. In addition, comparison of our 3D simulations with simplifications to 2D indicate that 2D models misrepresent critical aspects of the system. This has important implications for modelling of systems ranging from geothermal resources and mineralisation to carbonate diagenesis, including hydrothermal karstification and ore genesis as well as dolomitisation.