화학공학소재연구정보센터
AAPG Bulletin, Vol.94, No.6, 889-917, 2010
Reactive transport modeling of early burial dolomitization of carbonate platforms by geothermal convection
Reactive transport models (RTMs) permit quantitative investigation of diagenesis and its effects on reservoir quality. The RTM TOUGHREACT is used to investigate diagenesis in an isolated platform driven by geothermal (Kohout) convection of seawater, which has been invoked to explain dolomitization during early burial. Previous short (0.1 m.y.) RTM simulations suggested that convection can drive dolomitization, mostly at greater than 50 C, and anhydritization, but complete dolomitization requires greater than 30-60 m.y. Our more extended RTM simulations (<30 m.y.) indicate significant nonlinearities in the system, consistent with high-temperature experiments, with parts of the platform completely dolomitized within 10-15 my. As dolomitization proceeds, the process becomes predominantly flux controlled, with development of a wedge-shaped dolomite body, which thins from the margin to the interior, at considerably shallower depth and cooler temperatures (20-30 C) than suggested by short simulations. Dolomitization is relatively insensitive to boundary conditions such as relative sea level and platform geometry but is significantly slower in circular than elongate platforms. Sediment permeability and reactive surface area, commonly inversely related, are key controls. Dolomitization is limited to the margin of low-permeability muddy platforms despite a high reactive surface area. Dolomitization of more permeable grainy platforms is limited by a lower reactive surface area, occurring only in the platform core due to widespread cooling. Sedimentary layering produces a complex diagenetic stratigraphy, dolomitization favoring more reactive beds at shallow depth where permeability is not limiting, but switching to more permeable beds at depth. Bank-marginal fracturing limits dolomitization of the platform interior, whether the fractures are baffles or conduits for flow.