Scal validation

[petengr]

hello,

can some one give simple and straight forward practical step on how to validate a SCAL report. i read many theoretical info. i need practical ones.

thanks

[temr]

It is like to advise something on nothing
What do you want to validate ???

[petengr]

Hello temr,

I need to validate lab data of O-W relative permeability data, and capillary pressure data.

Regards

[temr]

So for sumulation purposes you will need 2 main thing
1 is sw vs cap pressure this normally uses in simulation to model saturation and transition zone
2 you will need to validate end points Swi and Sor vs poro or perm and Krwo Krow exponents , also Krowmax at Sw and Krwomax at Sor
3 the same thing for gas if you have 3 phase system
So i gave you guidelines now i would advise to read this topics in more detail
Use eclipse manual SCAL section

[Yengineer]

Unless there is a major mistake, you usually cannot validate laboratory data. If it is possible you can request lab crude datta and check them for inconsistencies (leaks on the rig, abnormal pressures etc) but this requires a good understanding of rig setup and the volumes.
Checking their data against eclipse usually does not bring anything. Laboratory crude data is feed into simulator and the curves are produced. Eclipse will not detect anything wrong with them, although they might be totally bullshit. As smd has said: shit in <-> shit out.

[temr]

Will agree with Yengineer

[petengr]

Thank you temr, and yengineer. Thank you for your help!

[DAH7542]

I'll add some simple checks on SCAL data prior to use this information in Eclipse, or other numerical simulators:

1. Check the base permeability of the Kr lab-data (usually Ko @ Swir). Most of the times, however, you'll need to input Kr data into the simulator using absolute permeability (Kabs) as the base permeability.
2. Smooth the Kr curves to prevent convergence-problems, these are associated with sudden changes in the slopes of the kr curves.
2. When 3-phase calculation is active in Eclipse (or other numerical simulators) this conditions must be satisfied: Krow @ Swir = Krog @ (Sg=0). Eclipse checks that this conditions is satisfied, if not, you'll get a problem.

[kader_007]

a simple and straight forward rule of thumb:
plot your points Kro, Krw vs saturation on a linear scale and semi-log scal too
=> the the trend should be concave upwards and smooth (no strange spots) for linear scale (as the theoritical shape let's say)
=> the same thing applies for semi-log plot (the only diffrence is the direction of the shape - concave downwards)
those are the simple points I do remember from SCAL course I did with Stiles...obviously a thorough QC should be taken.