relative permeability curves - Modified Corey Brooks

[itag]

I cant think of a specific article - but the steps are as follows assuming krw, kro curves.

1. modified corey asserts: krw ~ Ew*Swn^Nw , kro ~ Eo*Son^No where Ew, Eo are constants and Nw, No are the constant exponents. Swn Son are "normalised" saturations.

2. Swn = (Sw-Swir)/(1-Swir-Sorw) , Son = 1-Swn = (So-Sorw)/(1-Swir-Sorw)

3. Swir is the irreducible water sat. Sorw is the residual oil sat. ( that vary sample by sample)

4. For each of your samples your need to identify Swir, Sorw ( not as easy as you think - esp Sorw) then calculate the equivalent Swn values in the range [0,1]

5. Plot all the curves on krw, kro vs Swn * do they look the same ? ( this is where the fun begins) - do you average / discard / bin ??

6. Some folk normalise the y-axis too per sample Ew'=1=krw/krw_max=krw/Ew and plot krw' , kro' vs Swn ( I tend not to do this as there's more info in Ew than Ew' )

You will find that ALL lab data need to be corrected - and this is best done by iteratively fitting Nw, No, Sorw via excel until you get a good match. Corelabs report Sorw from the highest Sw achieved - this is usually not the 'real' Sorw. Similary Steady State coreflood produce different Sorw's than Unsteady State and centrifuge tests ( for good reasons). ( I recommend doing this is general - but this wasnt the question you asked). You will notice that I left Swir out - as this variable tends to be "better" defined by lab data .... its usually Sorw that is the focus.

I've noticed such questions arise time and time again - if you want to send me your raw data - I'll have a look and provide a step solution for the whole group that the admins can make a sticky ( if they want). Take all mention of the reservoir off the data sets.
My experience is that up to 50% of SCAL data is often "off trend" for various reasons - and the user needs to seach for the "jewell" that are left .... ( leave me a message about how to contact you)

The steps listed above are fairly easy to do ... but further refinements may help particulary if the 8 data sets come from very different poro-perm sets.

itag

[Shakespear]

From anihitas link you may also find this interesting

[link Point to another website Only the registered members can access]

[link Point to another website Only the registered members can access]