here is the region( horizontal inlet) where the viscosity ratio is more than 1e5. How can i improve the mesh quality of that specific area?

Ok, some thoughts:
1) Do you really need the small inlet pipe? This is numerically painful, because the length (and therefore also timescale) of the pipe is much smaller than the one of the tank. Can't you just imprint an inlet surface to the pipe wall and use this little surface as an inlet?

2)I wouldn't use a prism layer at the tank wall. You normally use them, when the flux is parallel to the wall (such as in pipes, at wings,...). Then, the main advantage is, that you can resolve the flow in one direction (perpendicual to the wall) and not too much in the others. Also some numerical errors cancel, when the flux is perpendicular to the prism surfaces. This is not the case in your setup: the flux is not strong along the tank surface! I don't see any sense in a refinement there. A refinement at the inlet and the region where the inlet flow flows along would be good!

semi circular flow
 

I have an orifice in the ground from where liquid pool starts coming.
The liquid pool is supposed to spread in a semicircular direction.
How to specify flow direction for this condition?
In the normal cartesian coordinate system the pool is to spread in negative and positive y direction and in positive x direction alone.
I am not able to specify (+1 & -1 for y axis) . Is there a way to do it?

reg LNG material properties
 

Can you tell how the properties (density , latent heat , boiling point) that are varying with temp were given as input in CFX.
What type of material was chosen (pure substance,mixture , refrigerant?)

water volume fraction created opposite direction
 

the water volume fraction is changed .it should have been horizontal.kindly help

Hi, i am also trying to understand the stratification of LNG in storage tank and facing the same problem that how to make geometry. So, can you please help me regarding this.