Investigation of partial water saturation effects on diffusion in shale

A new method for testing the effect of partially saturated conditions on aqueous diffusion was developed using samples from the Upper Ordovician Queenston Formation shale from the Michigan Basin of southwest Ontario, Canada. Effective diffusion coefficients (De) were determined for iodide tracer on duplicate cm-scale samples from a core segment. Partially saturated conditions were created with a new gas-ingrowth method that takes advantage of the variability of N2 solubility with pressure. The method is designed to create partially saturated pores, quantify the level of partial gas/brine saturation within the tracer-accessible pore space, and measure De under fully porewater-saturated and partially gas-saturated conditions for the same sample. X-ray radiography is used with an iodide tracer for quantifying the degree of partial saturation and measuring De. The saturated De values range from 2.8 × 10−12 to 3.1 × 10−12 m2/s. Following generation of a gas phase in the pores (average gas saturations of 4-6.7%), De values decrease by 20-22% relative to the porewater-saturated condition, indicating that the tortuosity factor (ratio of constrictivity to tortuosity) is sensitive to saturation. The data suggest that a relatively small fraction of the pore space dominates the solute transport. The gas-ingrowth method was successful for generating partial gas saturation, but the distribution of the gas phase is non-uniform, with relatively high gas saturations near boundaries and lower saturations in the interior of the samples.