Evolution of a brine‐saturated layer at the base of a ridge‐crest hydrothermal system

Abstract

Abstract. Vent fluids with chlorinities significantly greater than seawater are indicative of mixing between hydrothermal seawater and brines that have formed as a result of supercritical phase separation. We relate the evolution of a brine-saturated layer to magmatic and tectonic processes at the base of a ridge crest hydrothermal system. A conductive heat balance indicates that under steady state conditions such a brine-saturated layer is typically ~ 1-10 m thick and contains 10 7 to 10 8 kg of brine with 6 -10 times the chlorinity of seawater. Simple mixing and fractionation models indicate that the depletion and replenishment of brine-saturated layers may occur on a timescale of years to centuries. A timescale of years is consistent with temporal changes in chlorinity of fluids at vents in areas of recent, known dike injection. In these systems, brines may be undergoing depletion by mixing with liquid-phase seawater. Vents stably emitting high-chlorinity fluids on timescales of decades may represent older systems in which brines are being depleted by diffusion into an overlying liquid-phase hydrothermal system.

Depletion of a brine layer at the base of ridge-crest hydrothermal systems

“…The depletion times observed in our simulations are larger than those obtained from simple physical models of di¡usion and dispersion across the interface [10]. We attribute this di¡erence to the lower permeabilities used in our models.…”

“…The brine salinity may exceed 50 wt%, as found in £uid inclusions in ophiolites [3,9]. It is likely that the buoyant fresh water phase ascends quickly into the overlying (subcritical) £uids, while most of the dense brine stays within the supercritical area underneath [1,10,11]. The observed variations in vent chlorinities are well explained by mixing of hydrothermal seawater with either the (condensated) fresh water phase or its complementary brine [12,13].…”

“…We attribute this di¡erence to the lower permeabilities used in our models. Under hydrological parameters similar to those of Lowell and Germanovich [10], the £uid in the brine layer will certainly convect. From our experiments, this implies that the brine layer will not be stable and the density interface will vanish either by convective breakdown or upward migra- tion, on a time scale much shorter than the one predicted for depletion across a purely di¡usive/ dispersive interface [10].…”

“…Simple analytical models of the basic physics of brine depletion suggest that the generated brine layer has a lifetime of years to centuries, from the moment that the system has returned into the subcritical thermodynamical regime [10]. However, these methods do not take into account the £uid-dynamical interaction of the brine with the overlying liquids.…”

Cellular convection models of mid‐ocean ridge hydrothermal circulation and the temperatures of black smoker fluids

“…Numerical models suggest that double-diffusive systems may be unstable in low-porosity media [Rosenberg, 1991] and that phase-separated brines may be flushed out before large volumes accumulate. Simple physical models of the brine layer suggest that it may have a lifetime of years to centuries [Lowell and Germanovich, 1997] If the stable pattern of convection is single-layer circulation, then an alternative explanation must be found for the uniformity of maximum venting temperatures. Lister [1974,1983] argues that the maximum temperature of circulation is determined by the rigidus, the temperature at which rocks crack because ductile creep rates are insufficient to accommodate thermal contraction.…”