# Heterogeneity Driven Trapping at the Pore-Network Scale in Edwards Brown Dolomite

## TL;DR

This study examines how pore-scale heterogeneity in a dolomite rock affects CO2 trapping, showing that capillary barriers significantly influence fluid immobilization and plume migration.

## Contribution

The study provides new insights into how capillary barriers at the pore scale impact CO2 trapping in heterogeneous geological formations.

## Key findings

- A downstream low-porosity region acted as a capillary barrier, remaining brine-saturated during decane injection.
- Pore-network analysis revealed limited connectivity with over 30% of pores connected by two or fewer throats.
- Ganglia analysis showed elevated trapped volumes behind the barrier, indicating enhanced trapping and reduced accessible pore volume.

## Abstract

Trapping is a key control governing the stability and
long-term
containment of CO2 within geological storage formations,
with residual trapping at the pore scale being well established and
routinely incorporated into reservoir simulation models. In contrast,
integrating the effects of capillary trapping arising from spatial
variability in capillary entry pressure at the micron to centimeter
scale remains a challenge for field-scale models, despite clear evidence
of its influence on plume migration. Studying pore-scale heterogeneity
allows direct quantification of how heterogeneity in pore connectivity
and throat geometry translates into capillary entry pressures and
snap-off mechanisms, which ultimately control trapping efficiency
and is not often resolved at the continuum scale. In this study, we
performed flow experiments with brine and decane under capillary-dominated
conditions (C
a = 2.6 × 10–7) while acquiring time-resolved 3D micro-CT images at 5.6 μm
voxel size on a 12 mm by 60 mm rock sample. Fractional-flow drainage
and imbibition steps were imaged at steady state. Segmented volumes
were analyzed with pore-network analysis and trapped volumes were
investigated with ganglia volume and count analysis. The sample contains
a downstream low-porosity region that acts as a partial capillary
barrier. This region remained brine-saturated even during 100% decane
injection, indicating entry pressures above the applied capillary
driving force. Pore-network analysis showed limited connectivity where
the resolved coordination number is approximately only 2, with more
than 30% of pores connected by two or fewer throats. The relationship
between local initial and residual saturations shows that, within
the barrier region, the two values are nearly equivalent, indicating
negligible displacement of the mobile phase demonstrating minimal
displacement and enhanced trapping. The ganglia analysis shows that
the volume and count of ganglia trapped behind the barrier remained
elevated after imbibition. These results show that capillary barriers
increase immobilization while reducing accessible pore volume. This
has an influence on plume migration pathways and should be captured
in upscaled models for storage in heterogeneous formations.