When a fuel gas is released into a compartment it will mix with the available air under all possible actions of turbulent jet mixing, buoyancy, turbulent mixing with ventilation air and molecular diffusion. Molecular diffusion has shown to be extremely slow in comparison to the other effects and can be neglected in many applications.

Several experiments with gas leaking into rooms have been conducted at the British Midlands Research station as reported in reference ^[4]. An important conclusion drawn from these experiments is that, "for leaks of natural gas which occur near the floor, the time taken for uniform conditions to be established within the layer which is formed is short". Many experiments have shown this to be true. A layer is formed very quickly between the point of leakage and the ceiling. This layer's concentration gradually increases with time, until a steady state is established. (Note that this phenomenon isn't necessarily true for highly buoyant gases such as hydrogen)

The inverse effect occurs for a negatively buoyant gas such as propane. A layer of somewhat uniform concentration is formed very quickly between the point of leakage and the floor. Obviously, such results depend on the size of the room, the leak rate and the amount of air movement in the room.

For situations where such a "two layer" dispersion is valid, a simple mixing model can be used to estimate the bulk concentration of gas above or below a leak, depending on...