How to model hydrocarbon and hydrogen jet fires in Phast™/Safeti 8.6

Learn about recent jet fire modelling advancements in Phast

Welcome to our presentation on jet fire modelling of hydrocarbons and hydrogen. This how-to video is presented by Jan Stene, based on development and validation work performed by DNV for the new jet fire model tailored for hydrogen and syngas. The foundation of this new model lies on the work conducted by Derek Miller from Air Products.

 

The video discusses the context, methodology, and comparative analysis of the new Miller model against existing models and experimental results.

 

Highlights: 

 

Introduction  

  • Jet fire models are essential for predicting radiation from hydrocarbon fires.
  • Existing jet fire models rely on hydrocarbon data, which may not accurately predict radiation from low-luminosity flames like hydrogen.
  • The jet fire model developed by Derek Miller bridges this gap by addressing both hydrogen and syngas scenarios.

 

Jet fire models 

  • Jet fires vary in fuel composition, phase, release direction and environmental conditions. 
  • Accurate modelling requires considering these variations to properly assess potential impacts. 
  • Different models exist, ranging from complex Computational Fluid Dynamics (CFD) to simpler semi-empirical models. 

 

The Miller model 

  • The Miller model is an extension of the Chamberlain and Johnson models and uses semi-empirical data to predict hydrogen jet fires accurately.
  • Regarding flame shape, the Miller model represents the flame as two segments, a momentum-dominated part and a buoyancy/wind-dominated part. 
  • Regarding flame radiation, a multi-point source model distributes radiation along the flame's centre line, peaking at around two-thirds of the flame's length. 

 

Phast jet fire model updates based on the Miller model 

  • Phast has been adjusted to include lift-off distance, cross-wind effects, continuous radiation distribution, and planar observer calculations. 
  • These updates enhance the model's accuracy without significantly altering results from Miller's original formulation. 

 

Model validation 

  • Extensive comparisons between the Miller model and experimental data demonstrates improved accuracy compared to older models. 
  • For horizontal releases, the updated Miller model outperforms the Johnson model, providing better radiation predictions for hydrogen mixtures. 
  • For vertical releases, the Miller model surpasses the Chamberlain model in predicting vertical jet fires. 
  • While the Miller model is tailored for hydrogen, it also performs adequately for hydrocarbon jet fires, comparable to existing cone models.

 

Recommendations for model use 

  • For vapor jet fires with low luminosity gases, the new Miller model is recommended. 
  • For hydrocarbon fires, it is recommended to consider using the Johnson model for horizontal releases and the Chamberlain model otherwise. 
  • For two-phase and liquid jet fires, the Cook model is recommended. 

 

Limitations and further work 

  • The semi-empirical nature of the models limits their accuracy beyond the data range used during development. 
  • Future work will aim to better understand vertical jet fire data scatter, improve lift-off distance accuracy, and extend the model to two-phase hydrogen jet fires. 

 

Conclusions 

  • The new Miller model represents a significant advancement in modelling low-luminosity jet fires, offering improved accuracy and reliability. 
  • This model is included from Phast and Safeti release versions 8.6 and later. 

 

Acknowledgements 

  • Special thanks to Air Products and Derek Miller for their collaboration and data sharing, which have been instrumental in this work. 

 

For detailed information, please refer to the full video.