I'm modeling a transient hydrogen (1kg/sec) release within the wake of a building. After 650sec of simulation time a false source of hydrogen appears within my domain, away from the original plume... a few seconds after it reaches the outlet another source appears upwind of the building...

any ideas why?

THANKS TO ALL!

I ran into a similar problem years ago.

An error in mass conservation (poor convergence in the pressure correction equation) translates into a source (or sink) term in the conservation of one of the gas species being simulated. In your case, your hydrogen transport equation will react to an overall conservation error by showing a false source or sink of hyrdrogen in the same cell.

In my case, the solution was to tighten down on the convergence criteria for mass conservation in each cell. The false sources and sinks in the second component disappeared.

Of course, tight mass convergence can be very expensive computationally. Good luck!