recurring issues with mass/species conservation in fluent
 

Dear all,

There are 2 major recurring issues I am encountering in ANSYS FLUENT (i've worked with 14.5, 15 and 16), and I would like to know if more of you have experienced these issues, and perhaps have workarounds.

Problem 1 is the conservation of mass in multiphase simulations with a pressure-outlet boundary condition. I am modeling a tank/bubble column, with mass-flow or velocity inlet, and pressure outlet. There is a freeboard region filled with gas only (20% of the domain height), and the pressure outlet has a 100% gas backflow specified. There is no feed of the continuous liquid phase.
Still, the mass of liquid is not conserved; I've seen it both going up and down. This happens both in steady state and transient simulations. the imbalance happens at the outlet, but I cannot see why - especially I cannot understand why the water mass sometimes increases. In either case, the simulation does not converge, or converges with a large part of the water gone.

Problem 2 has to do with species tracking. I switched to a degassing B.C. (removing the freeboard region, so the tank is filled with gas-liquid mixture to the top), so that problem 1 is gone. However, when I enable the species equation, there is a continuous increase in the mass fraction of the non-bulk species. So, I feed gas with an oxygen content of 20%, but at the outlet boundary, gas with a 100% oxygen content flows back. I suspect this has to do with the backflow condition at the boundary - however, one cannot specify that for the degassing b.c.

Do these problems sound familiar? Any suggestions or workarounds?
Any comments would be of enormous help, since I am pretty stuck at this moment - and with the amount of work already done in FLUENT, switching to CFX would be a pretty unfortunate last resort.

So, I guess nobody ran in to these issues (or has a workaround?)