LNG dispersion

Accurate prediction of cryogenic gas dispersion, vapor cloud evolution, and safety distances following LNG releases, supporting risk assessment, design optimisation, and regulatory compliance

Technical capabilities in MPflow

MPflow provides advanced CFD capabilities for simulating LNG releases and cryogenic gas dispersion in realistic environments:

  • Simulation of multi-phase, multi-component flows including liquid-to-gas phase change
  • Modelling of cryogenic effects such as strong density gradients and thermal interactions
  • Accurate prediction of dense gas dispersion and gravity-driven spreading
  • Treatment of turbulence–buoyancy interaction in atmospheric conditions
  • Integration of terrain, obstacles, and industrial geometries
  • Support for transient releases, including spills, jets, and pool evaporation

MPflow solvers are fully compatible with OpenFOAM libraries and are designed for robustness in highly transient and strongly coupled thermodynamic conditions.

Our CFD approach

Using MPflow, MultiFluidX delivers high-fidelity simulations of LNG dispersion, capturing the full physics of release and cloud evolution:

  • Resolution of LNG vaporisation and pool spreading dynamics
  • Prediction of gas concentration fields and flammable cloud regions
  • Modelling of heat transfer between LNG, ground, and ambient air
  • Simulation of wind-driven dispersion and atmospheric stability effects
  • Accurate representation of dense gas slumping and stratification
  • Support for hazard distance estimation and safety zone definition

Physics-based ML acceleration

Our ML-CFD framework enhances LNG dispersion modelling by accelerating key processes while preserving physical fidelity:

  • Development of surrogate models for rapid dispersion prediction
  • ML-enhanced modelling of turbulence and mixing in atmospheric flows
  • Fast estimation of gas cloud footprint and concentration thresholds
  • Real-time evaluation of multiple release scenarios and environmental conditions
  • Hybrid workflows combining CFD accuracy with ML speed

Why ML-CFD matters for LNG Dispersion

  • 7000× faster predictions → near real-time LNG dispersion assessment
  • 5–7× acceleration with ML-enhanced turbulence models (RANS/LES)
  • <5% deviation from CFD in concentration and dispersion patterns
  • Instant multi-scenario analysis for safety and risk planning
  • Scales across weather, terrain, and release conditions
  • Ready for digital twins and real-time monitoring systems