Permeability is a critical parameter for the petroleum geologist. By simulating the processes of compaction and cementation in a model porous medium, we have gained a new understanding of how permeability is controlled in reservoir sandstones. This understanding can be used predictively for simple sandstones. The geometry of the model pore system is completely defined, so permeability can be calculated directly using a network model for flow. The calculation is based on first principles and is physically rigorous. In contrast to many previous efforts to predict permeability, there are no adjustable parameters in the calculations and no additional measurements or correlations (e.g., capillary pressure data or pore system data from thin sections) are required.

The model-derived porosity-permeability trend for a compacted or quartz cemented sandstone, or a sandstone having a combination of these processes, closely matches measurements on Fontainebleau sandstone samples whose permeabilities span nearly five orders of magnitude. The model also correctly predicts mercury injection measurements of pore throat size distribution for the Fontainebleau sandstone.

Pore-scale geometric features of the model are found to be spatially correlated, and this departure from randomness significantly affects macroscopic properties such as permeability. The agreement between predictions and measurements suggests that spatial correlation is inherent in granular porous media and consequently, uncorrelated (or arbitrarily correlated) models of transport in such media are unlikely to be physically representative. We also discuss extending the model to predict properties of more complicated rocks.