Laboratory and literature insights into uncertainty in CO2-water relative permeability at low water saturations

Abstract

Determining CO₂-water drainage relative permeability generally requires laboratory experiments, followed by numerical history matching. However, achieving low water saturations in the laboratory is challenging. Consequently, the relative permeability values at these low saturations—though essential for field-scale modelling—must be extrapolated, introducing significant uncertainty.

Previous studies used continuous mathematical functions—such as Corey or LET—to define relative permeability curves across the full saturation range. In such functions, changes to curve parameters affected both high and low saturation values, masking the specific uncertainty present at low saturations. In this study, we reanalyzed published data revealing a wide range of plausible relative permeability values at low water saturation, all of which yield equally good history matches—indicating substantial hidden uncertainty in this region.

To mitigate this, we performed laboratory experiments with extended injection of water-saturated CO₂ to 78 pore volumes (PV), achieving 34 % water saturations, much lower than commonly reported. Following this, desaturation was performed at constant pressure using a porous plate to further reduce water saturation to 0.225, enabling direct measurement of maximum CO₂ relative permeability. Results indicate that extending CO₂ injection reduces uncertainty in relative permeability at lower saturations, though experimental limitations persist below 0.34 water saturation. Including porous plate data significantly improves reliability by applying higher capillary pressures representative of field conditions.

This work highlights the necessity of advanced experimental designs to extend the reliability of CO₂-water relative permeability measurements to lower water saturations. These findings are crucial for enhancing predictive accuracy in field-scale CO₂ sequestration modelling.