Simple models of CO2 release from metacarbonates with implications for interpretation of directions and magnitudes of fluid flow in the deep crust

Simple one-dimensional models of coupled advection-hydrodynamic dispersion-reaction are used to investigate processes of CO2 release from metacarbonate beds during deep crustal (similar to 8 kbar) Acadian prograde metamorphism in New England, USA. Two broad models in which reaction progress is controlled by gradients in H2O-CO2 fluid composition between different rock types are presented. In the first, diffusional exchange of volatiles across lithologic contacts is significant. CO2 generated during prograde temperature (T) rise is transported away from metacarbonate layers to surrounding (1) metapelitic layers which generate H2O by dehydration and/or (2) flow conduits (e.g. permeable layers or fractures) for externally derived, elevated X-H2O/X-CO2 fluids. H2O is transported from the surroundings into the metacarbonate layers and drives further mineral reaction. In the second model, reaction in metacarbonate layers is driven mostly by layer-parallel flow of external fluids with elevated X-H2O/X-CO2 derived from, for example, dehydrating schists or outgassing magmas. For both models, the X-CO2 of the fluid within metacarbonate layers is generally predicted to increase with increasing grade from the Ankerite-Oligoclase to the Amphibole zones, and then decrease in the Diopside zone-key relationships that are commonly observed in the field. The slow reaction progress in metacarbonates driven by progressive dehydration of surrounding metapelite from greenschist to amphibolite facies probably requires time scales of fluid-rock interaction comparable to the duration of the Acadian orogeny (similar to 10(6)-10(7) my). Intense episodes of fluid flow through conduits such as fractures may produce veins and alteration selvages over fluid-rock interaction times as short as 10(3)-10(4) years. Model results emphasize that the accuracy of field-based fluid flux estimates depends critically on correct identification of mass transport processes. The modeling suggests that reactive transport of volatiles between metacarbonate layers and their lithologically heterogeneous surroundings can account for basic T-fluid composition-reaction progress relationships observed in much of the Acadian orogen of New England. The results provide an alternative to up-T flow scenarios that account for these relationships by large, pervasive, horizontal fluid fluxes up regional T gradients.