The Upper Devonian Grosmont platform in the Western Canada sedimentary basin is a pervasively dolomitized giant heavy-oil reservoir with reserves of 317 billion bbl of bitumen. The principal type of Grosmont platform dolomite formed early and on the basis of stratigraphic and geochemical evidence is interpreted as early diagenetic reflux dolomite.
We use a numerical ground-water flow model to investigate the viability of reflux to dolomitize the Grosmont platform. We simulate reflux at four key stages of platform evolution, incorporating the transient effects of changes in platform architecture, rock properties, and the salinity of platform-top waters.
The pattern and magnitude of reflux is critically controlled by permeability and the distribution of platform-top brines, which are concentrated up to gypsum saturation. Reflux flow is focused in the relatively permeable carbonates of the Grosmont Formation and is from the platform interior toward the platform margin. The 120-m-thick shales of the Ireton Formation that separate the Grosmont and Cooking Lake formations restrict cross-formational flow and brine transport. During a 100-k.y. period of relative sea level rise and platform-top drowning, brines of reflux origin continue to sink and entrain platform-top waters (latent reflux). Where the intervening aquitards are thin or absent, reefs of the Leduc Formation capture reflux brines from the overlying Grosmont platform and focus cross-formational brine transport. Lateral contrasts in salinity are sufficient to drive a series of free convection cells in the relatively permeable reefs of the Leduc Formation.
Computed distributions of fluid flux in conjunction with magnesium mass-balance calculations that incorporate the range of uncertainty, particularly in permeability, support the suggestion that the reflux of gypsum-saturated brines could have formed much if not most of the dolomite in the Grosmont Formation in the 1.6 m.y. available.