trapping effect in relative permeability calculation of two phase?

[Gal-Tzvik]

Following the proposed algorithm given in the examples, I witnessed that this does not take in to account the effect of the trapping within the invasion. After trying to add the trapping effect there seemed to be some discrepancies between the drainage results and the relative permeability limits of calculation. I was wondering if maybe there is something wrong with my code/am i not looking at this properly considering he evaluation of two different properties, or is this due to the inherent properties of the library functions.

As an example, you can see my drainage curve with and without trapping, where the non wetting phases invades only ~80% percent of the total volume when considering trapping but when calculating the relative permeability the invasion sequence only reaches ~55%.

Drainage Curve

No trapping - RP

RP with trapping

[jgostick]

These should both be using the same invasion sequence numbers, so not sure what the discrepancy is, but my first guess would be that the x-axis in the rel perm curve is not including the volume of throats, only of the pores.

[Gal-Tzvik]

I am aware of this, therefore i dont understand why i have differences in a general characteristic such as the maximum non wetting phase sat.

[jgostick]

As you mentioned, the inlet's are treated differently in each case, which may lead to subtle changes in the pattern, but I would not expect this to be huge (assuming the network is large). Regarding my other point about the differences in trapping between invasion vs drainage, this paper has some more discussion in terms of image-based invasion. The point is that drainage will always under-estimate trapping since it only evaluates it after many pores are filled, whereas invasion evaluates is after each pore, so catches every trapping event.

[Gal-Tzvik]

Thank you i will definitely look into it. But just so i understand what you mean, the findings considering trapping with the ip algorithm should present more accurate results?

[jgostick]

The IP is more rigorous when looking at trapping, but this is only from a numerical standpoint. In physical terms trapping is a bit more complicated since the trapped phase can evaporate or move around via corner flow. So while IP does a better job at finding trapped defending phase, I can't say it's more or less accurate compared to reality.

[Gal-Tzvik]

Thank you very much for the support!