CO2 Leak in CCS: OpenFOAM CFD Reveals Lethal Zones in Low Elevations

Published on May 05, 2026 | Translated from Spanish

A massive carbon dioxide leak in a Carbon Capture and Storage (CCS) system has exposed a critical design flaw in a safety valve. The incident, which could have caused a catastrophic asphyxiation event, has been analyzed using 3D CFD simulation with OpenFOAM. The results show how CO2, denser than air, dangerously accumulates in low-lying areas, creating pockets of lethal gas that the original design did not account for. ๐Ÿ’€

3D CFD simulation of massive CO2 leak in CCS facility, showing lethal accumulation in low-lying areas

Modeling and simulation of the CO2 dispersion scenario ๐Ÿงช

To reconstruct the accident, a digital twin of the plant was created using AutoCAD Plant 3D, where the pipeline and the faulty valve were modeled. The real geometry of the environment was captured through photogrammetry with Bentley ContextCapture, integrating point clouds of the terrain and surrounding structures. With OpenFOAM, the Navier-Stokes equations for compressible flow were solved, using a k-epsilon turbulence model and a refined mesh in areas of potential accumulation. The transient simulation of 600 seconds revealed that at ground level, CO2 concentrations exceeded 10% by volume in just 3 minutes, a threshold considered immediately dangerous to human life.

Visualization and lessons for industrial safety ๐Ÿ›ก๏ธ

The risk visualization was carried out in Lumion, where the CO2 concentration fields were integrated onto the realistic 3D model of the plant. The resulting animation clearly shows how the gas moves like a viscous fluid, occupying trenches, basements, and terrain depressions. This analysis demonstrates that CCS systems require a redesign of valves with redundancy and concentration sensors in low-lying areas, as well as forced ventilation systems. CFD simulation not only explains the failure but also becomes a mandatory tool for safety certification in carbon capture facilities.

As a CFD engineer, what key parameters should I adjust in OpenFOAM to accurately model the lethal dispersion of CO2 in low-lying areas during a leak in a CCS system, and how can I validate those results with experimental data?

(PS: Simulating catastrophes is fun until your computer melts down and you are the catastrophe.)