Dear vinomarky,

Thank you for your great comments. I have a very important conceptual problem.

The water saturation resulting from the log are the initial water saturation. When you use SWL instead of SWATINIT to import them to eclipse, your implicitly tell eclipse to scale the relative permeability curves. For example if in a cell the Sw of the log is 0.8, the SWL of that cell is 0.8. It means the water won't flow until its saturation becomes greater than 0.8. Isn't it unrealistic? Generally this method yields to very little water production (Since you impose immobility in each cell be setting SWL=Sw) which may not
be the case.

Why don't we use SWATINIT which honors the initial water saturation and changes Pc to match it?