Hydrogen Consequence Software Modelling – Not a Set-and-Forget Exercise

As discussed in previous articles, hydrogen releases do not behave the same way as hydrocarbon releases and are governed by a different set of physical principles. For example, the Chamberlain equation, widely used for methane jet fire flame length estimations, significantly overestimates flame lengths for hydrogen. This discrepancy arises because the Chamberlain equation was never developed for hydrogen and does not account for critical factors such as hydrogen's barrel release geometry, buoyancy effects, flame blow-off, pressure oscillations, etc.

Many commercially available consequence modelling software packages (no names provided here) are built around the Chamberlain equation or the guidance given in API 521. This creates potential pitfalls for the safety engineer who relies solely on these tools for fire consequence predictions. Similar concerns arise when modelling explosion consequences, where some software packages rely on legacy models such as TNO Multi-Energy model, Kinsella, BST curves, or the TNT equivalency model. It is worth stating clearly: the TNT equivalency model should not be used for hydrogen explosions, including physical (non-combustion) explosions.

Some software vendors have implemented workarounds or adjusted input parameters to better approximate hydrogen behaviour, and while these adaptations can bring results closer to reality, discrepancies can still occur—sometimes significantly—depending on what is being modelled. Results could range from relatively accurate to entirely misleading. Even dedicated hydrogen-specific modelling tools available in the public domain can yield considerably different outcomes from each other. This variability is not a reflection of poor modelling—both have been developed by reputable organisations—it does however highlight that, as with any tool, the quality of the outcome depends heavily on the knowledge and skill of the user.

What about CFD (Computational Fluid Dynamics)? CFD can indeed provide more realistic and detailed results for hydrogen release and consequence modelling. However, the reliability of CFD outputs is still influenced by the inherent limitations of the model, including mesh size, solver capabilities, calibration of the underlying equations, and, critically, the user's level of expertise.

So, how should a hydrogen safety engineer approach this?
 The key is preparation and understanding. Hydrogen safety engineers must:

·      Stay up to date with the latest research and experimental results on hydrogen behaviour.

·      Perform hand calculations where possible to develop a “ballpark” sense of expected outcomes.

·      Understand the physical theories and phenomena dictating hydrogen behaviour during release events.

·      Recognise the strengths and limitations of each available software package.

·      Select software and modelling approaches on a case-by-case basis depending on the specifics of each release scenario.

Unlike the hydrocarbon sector, where fluid behaviour upon release is well understood and software tools have been refined over decades, the hydrogen industry is still developing its empirical base. Every variation in parameters—pipe size, operating pressure and temperature, release height, orientation, and more—can significantly affect hydrogen release behaviour.

In summary:
Do not rely blindly on any one software package. Always have an independent understanding of what the consequences and magnitudes should be. Select the most appropriate software or method for each specific scenario—even if that means using different packages for different aspects of the analysis.

In consequence modelling for hydrogen, as elsewhere, it remains very much a case of horses for courses. Don’t be swayed by colourful 3D simulations or graphs. They may be pretty but are they accurate? I’ll leave that for you to decide…