There is limited published research on the effect of water sprays on hydrogen deflagrations and deflagration-to-detonation transition, and more extensive data on water spray effects on hydrocarbon gas explosions. The results show the benefits, where there are benefits, to be highly scenario dependent. For example, Carlson et al. (Atomics International report, 1973) described hydrogen detonation tube testing with and without a 500-micrometer drop diameter water spray. They found the water spray prevented detonations at hydrogen concentrations of 20% to 24%, but not at 28%. Drop size is critical since the benefits of sprays depend on the droplets breaking up and partially vaporizing in advance of the propagating flame front. If the droplets do not break up, they induce turbulence effects that can significantly increase flame speeds and exacerbate hydrogen deflagrations. Both the detonation ability and deflagration scenarios are discussed in Technical Aspects of Hydrogen Control and Combustion in Severe Light-Water Reactor Accidents, National Academy Press, 1987. 

The effects of water deluge nozzle spray on methane and propane deflagrations in vented simulated offshore platform enclosures were investigated in Phase 2 (1998 report) of a European test program called Blast and Fire Engineering for Topside Structures. Comparisons of test results in highly obstructed enclosures with and without the deluge spray showed lower pressures with the deluge spray. However, the sprays did increase the initial flame speeds, which can be either good, if the flame induced velocities are large enough to cause droplet breakup, or bad, if the droplets remain intact. The authors conclude that deluge sprays prevent runaway flame acceleration in situations with inherently high flame speeds but will not be effective in applications where only low flame speeds are generated. Thus, the spray would be expected to be effective for near-stoichiometric H2-air mixtures, but detrimental for concentrations in the range of 8% to perhaps 13%.

There are two parts for such a system to be effective. First, the system would have to activate quickly enough to establish a water mist throughout the region of interest (i.e., region occupied by a flammable gas mixture) before it could be ignited. This is challenging in terms of timing, and the impact of spraying water inside an enclosure filled with equipment not designed to get wet can be an issue. Second, such a system has to provide a sufficient density of water droplets in the right size range to have the desired effect. For conventional vapor cloud explosions (e.g., large vapor cloud occupying a portion of a process unit at a refinery or chemical plant), experimental work has examined the effect of water sprays. Getting a high enough density of the right sized water droplets has been shown to be effective at slowing the flame front, and hence reducing the resulting blast load. The goal has not been to put the flame out (i.e., quench the flame), as this would be much more difficult, but rather to absorb some of the combustion energy and hence slow the flame. 

The same type of research has been done for mine explosions in the U.S. For a hydrogen-air mixture inside an enclosure, it is theoretically possible to use a water spray (from a fast-acting system triggered on detection) to slow down the flame. This could reduce the demand on an explosion venting system or reduce the structural capacity required for a given vent system. There are a number of recent publications on this topic (i.e., slowing the flame front propagation velocity, reducing the maximum blast pressure, and/or extinguishing the flame for hydrogen or mixtures containing hydrogen). However, the HSP is not aware of any such system for hydrogen explosion prevention or mitigation in industrial applications being available for purchase.