METAMORPHIC FLUID FLOW AT DEPTHS BEYOND THE BRITTLE-DUCTILE TRANSITION

The brittle-ductile transition is thought to define the depth limit for deformation by frictional sliding. Above this depth tectonic deformation may create permeable fracture systems that permit drainage of metamorphic fluids with negligible fluid overpressure. In this regime, fluid flow is independent of the stress state of the rock matrix and weak perturbations induced by topography or fluid density variations may give rise to complex flow patterns. In contrast, beneath the brittle-ductile transition the dilational deformation necessary to create fluid pathways can only be accomplished by fluid pressures that exceed the confining pressure. Although, this dilational deformation may be manifest by hydrofracturing, ductile compaction is responsible for generating the high fluid pressures and therefore determines the length and time scales for fluid flow. Association of fluid flow with compaction implies that beneath the brittle-ductile transition flow direction is dictated by the mean stress gradient. The paradox posed by the conditions required to maintain high fluid pressure with evidence for significant lower crustal rock strength can be explained if flow is accomplished by self-propagating high-permeability domains. Such domains would behave as weak inclusions imbedded within a strong matrix, i.e., adjacent relatively fluid-poor rock. Because the mean stress gradient within such an inclusion depends on both inclusion shape and orientation with respect to far-field stress both geometry and flow are sensitive to tectonic forcing. For compressional settings, this model has the counterintuitive implication that flow may be directed downward beneath the brittle-ductile transition to a depth of tectonically-induced neutral buoyancy. The depth of neutral buoyancy would act also act as a barrier to upward directed flow within vertically elongated domains, but would not hinder the propagation of sill-like domains. This behavior suggests an antagonistic relationship between the hydraulic potential and conductivity, such that conductive vertically oriented structural features would tend to evolve to less-conductive horizontal structures beneath the brittle-ductile transition, a conclusion supported by the geometry and depth of numerous mid-crustal seismic reflectors.