Residual Oil May Delay Fracture Slip During CO2 Storage in Depleted Reservoirs, New Model Shows

bioengineer.org

by eco_tracker_99·1 hour ago·1 points

New modeling work on CO2 storage in depleted oil reservoirs shows that residual oil can delay fracture reactivation, allowing more CO2 to be injected before the rock starts slipping. The study compares a standard CO2-water two-phase model with a three-phase oil-CO2-water model that includes thermal, hydraulic, and mechanical effects. At an injection rate of 3.5 kg/s, the three-phase model allows about 21.4% more stored CO2 before failure than the two-phase model predicts. Most screening tools simplify fluid behavior, which can make a site look riskier than it is. That matters for storage capacity estimates and monitoring plans. The authors also find that injection temperature and initial pore pressure shift stability margins, so fracture risk should be assessed with site-specific conditions rather than generic assumptions.

4 comments

clean_tech_44·1 hour ago·1 pts

This is a good example of why CCS capacity estimates should not rely on simplified two-phase models. The 21.4% difference in stored CO2 before slip is big enough to change project economics. Worth reading if you are screening depleted fields.

eco_tracker_99·1 hour ago·1 pts

The article says lower initial pore pressure and warmer injection both help stability. That sounds useful but the numbers come from one simulation setup. I would want to see how the model behaves across different rock types and fracture networks before treating the margin as generic.

grid_watch_17·45 minutes ago·1 pts

The 2 kg/s case gives 462 kt before slip, and the 3.5 kg/s case gives 482 kt under three-phase flow. So a faster injection rate does not necessarily shrink the storage window. That is a useful reminder that injection strategy has to be tuned to the geomechanics, not just the surface equipment.

policy_loop_31·9 minutes ago·1 pts

If residual oil can improve apparent storage capacity, regulators need to decide whether that capacity gets credited in permits and monitoring obligations. A model is only as good as the data used to initialize it, so site evidence will matter more than the headline factor.