[mbane]

I have convergence and results problems in my simulation. I am trying to simulate a 150 m steam pipeline in order to determine energy losses. The transport fluid is dry steam (water ideal gas) with 180C and 10 m/s at inlet. Outlet pressure is set to 9.8 bar, while reference pressure is 0 bar. I use a rough wall, with Heat Transfer Coefficient = 10 [W m^-2 K^-1], outside temperature 20C and included viscous term. The overall mesh size is 7000000 elements. Boundary layer has 15 layers with total thickness of 1.5 mm and grow rate of 1.1. The mesh was refined on all the pipeline curves. I am using the k-e model, high resolution for turbulence numerics and advection control with double precission.
The simulation converges just fine with low values of outlet pressure (I used the outlet pressure of 0 bar) but the results are unrealistic. The steam was cooled down for 30C which is impossible.
When I use the real value of outlet pressure (9.8 bar) the solution fails to converge below RMS 10-4.
I have attached the convergence diagrams.
Note that for all monitors other than h energy domain imbalance is 0%. For h energy it is 0.001%.
The results are again unrealistic because temperature drop between inlet and outlet is now 18C which is huge for a given problem.

Does anybody knows what might be a problem with my setup?

Thank you in advance.