A Case for Caution When Using Geochemical Models to Make Predictions about Dolomite

Dolomite reservoirs are common throughout the world and contain a significant portion of the known global hydrocarbon reserves. Despite centuries of research, why, where, and how dolomite forms in the geological environment is still unclear. In the past decade geochemical reactive transport models (RTM) have been used to test hypotheses about the hydrological settings and physicochemical conditions where dolomite forms. One potential drawback has been that model results rely on extrapolation of thermodynamic and kinetic data generated at temperatures higher than occur in many of the flow systems simulated. The present study highlights a common geological scenario where theoretical and model predictions are at odds with empirical evidence. Dozens of new high-temperature dolomitization experiments involving either crushed Iceland spar calcite or aragonite ooids in Mg-Ca-Cl solutions were conducted to test the effects of elevated salinity (NaCl and KCl concentrations) on the rate of dolomitization. Whereas theoretical predictions and 0-D (batch reaction) geochemical models created in TOUGHREACT using a number of thermodynamic databases predict that the effect of increasing salinity is a higher saturation index (SI) for dolomite, the high-temperature experiments produce less dolomite from more saline fluids. Interestingly, the decrease in the rate of dolomitization in the high-temperature experiments is accompanied by an increase in the magnesium content (mole% MgCO3) of the dolomite. Dolomite products produced in higher salinity experiments are also characterized by a higher degree of cation ordering. These observations suggest that the slower kinetics permit more magnesium to incorporate into the growing dolomite, and perhaps more importantly, into the correct cation positions within the crystal lattice. Considering that most fluids implicated in dolomitization are characterized by varying degrees of salinity, the experimental findings provide a simple example of why it may be wise to exercise caution when using the results of numerically-based geochemical models to predict how and where dolomite forms. This study also stresses the importance of how building a more robust empirical database will ultimately help improve geochemical models and thus our predictive capabilities