backbutton
Back to news section

Published 20.08.2022.

Comment on CGA G-5.5 standard and design advice for hydrogen vent masts

HYEX Safety supports clients with safety advice and studies for vent masts for hydrogen systems. Recently different clients have referred to formulations in “CGA G-5.5 standard for hydrogen vent systems” requiring vent mast L/D below 60 to eliminate the risk for deflagration (=explosion).

We were not familiar with the CGA G-5.5 standard, is it really saying this?

L/D = 60 is not a deflagration limit but has been referred to as a detonation limit for smooth pipes (typical DDT run-up distance). For hydrogen-air mixtures DDT run-up distances in smooth pipes are shorter than this. Knystautas and Lee reported experimental DDT run-up distances of L/D 40 to 50 in smooth pipes of D=174 mm and 520 mm, typical vent pipe diameters. For vent pipes potentially filled with reactive mixtures of hydrogen-air, severe detonations may be feared with flame speeds near 2000 m/s and local flame front pressures of 15 bar increasing to 30-50 bar or more at reflections.

Can detonations be prevented by ensuring L/D less than 60 or 40? Definitely not!

Vent systems of real applications are seldom straight smooth pipes. They have bends, junctions, expansions or more. Once a reactive hydrogen flame has passed a couple of bends (=obstructions) detonation run-up distances shorten dramatically, see slide from Dorofeev below.

Slide from presentation by S. Dorofeev at ICHS 2 (2007)

HYEX has performed many safety studies of hydrogen venting systems for maritime and land-based industry applications, including LH2, with CFD modelling of vent flows, mixing, plumes, ignition, explosion (deflagration, DDT, detonation) and purging processes experiencing lots of “learning-by-doing”. Strong explosions are definitely not prevented by “rules” like L/D < 60 but rather by understanding how explosive mixtures may form and be prevented inside the vent pipes. From our experience the main concern is not ignition and explosion at the onset of hydrogen venting before air is pushed out of the vent stack. Such incidents may be the more likely but usually have limited consequences. Our calculations indicate that the by far more severe scenario may develop long after the hydrogen is vented as air is slowly diffusing into the hydrogen filled stack. This must be prevented by dedicated purge procedures.

Finally, some observations highly applicable to hydrogen systems design:

  • Rules, Codes & Standards describing safe design are hard to make, not least for hydrogen.
  • Rules-based design solutions can be costly, inefficient, and unsafe.
  • Standard work is often unpaid and takes years. Do we manage to ensure balance between commercial interests and scientific content?