Skip to content

Performance Evaluation of Empirical Models for Vented Lean Hydrogen Explosions


Explosion venting is a method commonly used to prevent or minimize damage to an enclosure caused by an accidental explosion. An estimate of the maximum overpressure generated though explosion is an important parameter in the design of the vents. Various engineering models (Bauwens et al., 2012, Molkov and Bragin, 2015) and European (EN 14994 ) and USA standards (NFPA 68) are available to predict such overpressure. In this study, their performance is evaluated using a number of published experiments. Comparison of pressure predictions from various models have also been carried out for the recent experiments conducted by GexCon using a 20 feet ISO container. The results show that the model of Bauwens et al. (2012) predicts well for hydrogen concentration between 16% and 21% and in the presence of obstacles. The model of Molkov et al. (2015) is found to work well for hydrogen concentrations between 10% and 30% without obstacles. In the presence of obstacles, as no guidelines are given to set the coefficient for obstacles in the model, it was necessary to tune the coefficient to match the experimental data. The predictions of the formulas in NFPA 68 show a large scatter across different tests. The current version of both EN 14994 and NFPA 68 are found to have very limited range of applicability and can hardly be used for vent sizing of hydrogen-air deflagrations. Overall, the accuracy of all the engineering models was found to be limited. Some recommendations concerning their applicability will be given for vented lean-hydrogen explosion concentrations of interest to practical applications.

Related subjects: Safety
Countries: United Kingdom

Article metrics loading...



Performance evaluation of empirical models for vented lean hydrogen explosions

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error