Open knubo-equinor opened 3 years ago
The first issue should probably be solved elsewhere, as it is hard for pyscal to correctly error hard on input that gives extreme output. Emitting a warning has probably no effect (?). Pyscal only handles numerical values, it does not know which units are in use.
For the second issue, I don't get precisely what you mean. Example input and output?
I have tried to illustrate in the pptx enclosed. The issue is that for Sw=1 the Pc should be zero (but not always).
Pc-curves get scaled and what looks like a negligible entry pressure becomes large after initialization.
Regarding the first issue: you are probably right that a warning among zilion other warnings in Eclipse and FMU will not make an impression on the users. However, you have some assumption about units since you use gravity in SI units, as a standard- or?
The simplified J-function does not yield zero at sw=1
; https://equinor.github.io/pyscal/pyscal/pyscal.wateroil.html#pyscal.wateroil.WaterOil.add_simple_J
The documentation of the add_simple_J()
talks about which units are recommended to use, but the code itself does not care. So if you know what you are doing, you could utilize that to your benefit.
Should this issue be kept open?
When I open simulation projects, especially in FMU settings, that has used pyscal to generate saturation table the Pc-curves is often quite unrealistic. and contains large entry pressures. This seems to have become a company challenge.
The first issue is maybe a user fault, but isn't it possible to create a test so that PC doesn't go to hundreds of bars.
The second issue is that Pc <> 0 bar @ Sw=1. There might be entry pressures in the system but often its just a result of automatic workflow.