Effect of confining pressure on diffusion coefficients in clay-rich, low-permeability sedimentary rocks.

The effect of confining pressure (CP) on the diffusion of tritiated-water (HTO) and iodide (I) tracers through Ordovician rocks from the Michigan Basin, southwestern Ontario, Canada, and Opalinus Clay from Schlattingen, Switzerland was investigated in laboratory experiments. Four samples representing different formations and lithologies in the Michigan Basin were studied: Queenston Formation shale, Georgian Bay Formation shale, Cobourg Formation limestone and Cobourg Formation argillaceous limestone. Estimated in situ vertical stresses at the depths from which the samples were retrieved range from 12.0 to 17.4MPa (Michigan Basin) and from 21 to 23MPa (Opalinus Clay). Effective diffusion coefficients (D) were determined in through-diffusion experiments. With HTO tracer, applying CP resulted in decreases in D of 12.5% for the Queenston Formation shale (CP=12MPa), 30% for the Georgian Bay Formation shale (15MPa), 34% for the Cobourg Formation limestone (17.4MPa), 31% for the Cobourg Formation argillaceous limestone (17.4MPa) and 43-46% for the Opalinus Clay (15MPa). Decreases in D were larger for the I tracer: 13.8% for the Queenston shale, 42% for the Georgian Bay shale, 50% for the Cobourg Formation limestone, 55% for the Cobourg Formation argillaceous limestone and 63-68% for the Opalinus Clay. The tracer-specific nature of the response is attributed to an increasing influence of anion exclusion as the pore size decreases at higher CP. Results from the shales (including Opalinus Clay) indicate that the pressure effect on D can be represented by a linear relationship between D and ln(CP), which provides valuable predictive capability. The nonlinearity results in a relatively small change in D at high CP, suggesting that it is not necessary to apply the exact in situ pressure conditions in order to obtain a good estimate of the in situ diffusion coefficient. Most importantly, the CP effect on shale is reversible (±12%) suggesting that, for argillaceous rocks, it is possible to obtain D values that are representative of the in-situ condition by conducting measurements on re-pressurized samples that were obtained with standard drilling practices. This may not be the case for brittle rock samples as the results from limestone suggest that irreversible damage occurred during the pressure cycling.