Delayed ignition of high-pressure hydrogen releases – Experiments and engineering models

Abstract

Before operating its first hydrogen transmission pipelines, GRTgaz is developing methods and engineering models for risk assessment and consequences analysis. In this framework, GRTgaz decided in 2022 to launch an experimental campaign to understand the consequences of delayed ignition of high-pressure pure hydrogen releases, in open field. The tests also included ignited releases of methane-hydrogen mixture, with 2% and 20% of hydrogen. GRTgaz was associated with Storengy to fund this campaign performed by Gexcon AS. The tests took place at the test facility of Gexcon in Sotra island, close to Bergen (Norway).

The campaign was divided in two test series. The phase 1 aimed at characterising gas concentration in the release axis to provide validation data for simple models and to facilitate the positioning of igniter for the explosion tests, in phase 2. In this second phase, the releases were ignited by chemical devices or electrical sparks, located in the axis at different equivalence ratio. The release system designed by Gexcon enabled to perform 40 barg releases through calibrated orifices of 4 and 6 mm. In total, 15 tests were performed for the first phase with unignited releases and 29 explosion tests for the second phase. All tests are steady-state horizontal releases.

The test results provide comprehensive data to better understand hydrogen jet explosion and challenge engineering models. The maximal overpressures were well above the blind predictions, with records over 650 mbar close to ignition region, where about 200 mbar were expected. High-speed videos showed a tremendous acceleration around the ignition source that cause these intense overpressures coupled with very short-duration positive pulses. These results tend to indicate that the overpressure is produced in a limited volume compared to the flammable plume.

Then, simulations with the PERSEE + software are compared to the experiments. Dispersion results are in good agreement with the recorded concentration with a relative deviation around ± 30%. For overpressures, the prediction in the far field is also acceptable, whereas effects in the near field are often underpredicted. The article finally discusses about improvement of engineering models to estimate consequences of a delayed ignition of high-pressure hydrogen releases.