An experimental investigation into the ignition of leaking hydrogen

Abstract

The intent of this work was to determine how close an ignition source must be to a hydrogen leak to cause ignition of the leak. This distance was then compared to the maximum distance at which 4% hydrogen concentration could be found in the leaking hydrogen plume. This was done by constructing a device to produce a measured 566 l/min (20 SCFM) leak horizontally through the center of an aluminum clad 8 ft by 8 ft wall. The pure hydrogen leak was allowed to continue until it produced a relatively steady-state plume. The concentration of hydrogen was measured at various locations and compared to a computational fluid dynamics model of the plume. Electric arcs, and in some instances open flames, were utilized at various locations to determine those locations that would produce full ignition of the leak. The experimentally determined location, farthest from the wall, at which the leak could be ignited, was then compared to the computational and experimentally determined location of 4% hydrogen–air mixture in the plume. It was found that the maximum horizontal distance, from the horizontal hydrogen leak at Mach = 0.10 , at which the leak could be ignited was 57 in which was significantly closer than the location that contained 4% hydrogen (77 in). The maximum horizontal distance to ignite a Mach = 0.20 leak was 47 in which was significantly closer than the location that contained 4% hydrogen (88 in). Additionally, the ability of a 9.0% hydrogen in air flame to ignite paper was tested. One minute exposure to the flame did not ignite the paper.