Analyzing capillary pressure (Pc) is one of the most common methods to create petrophysical groups in carbonates, since the Pc explains the moment when the fluid starts moving in the porous media, how easy or hard is the displacement of the fluid and when there is no further displacement (irreducible saturation of the porous media).

In comparison to routine core analysis (RCA) data, MICP dataset is rather small. However, utilizing all the spectrum of data is critical, more specifically in highly heterogeneous carbonates, hence the need to populate petrophysical groups from MICP to RCA space.

To determine the petrophysical groups from the MICP the following parameters were used: entry pressure, normalized porosity, permeability, rock quality index, and the hyperbolic tangent of each to the MICP samples. The petrophysical groups generated by this method give a very discrete clustering where the ranges of porosity and permeability are well defined. For distributing petrophysical groups from the MICP domain into the RCA domain, a Bayesian Inference approach in the porosity permeability space was used. Porosity and permeability standard deviation and mean for each of the petrophysical groups, were used to build a probability density functions (PDF) which will be taken as our probable scenario to feed the Bayesian Algorithm.

The results from 350 MICP samples show that we could establish a reliable petrophysical groups distribution over 7000 RCA samples. In a second stage, petrophysical groups populated in the RCA space were used to train our logs and create a continuous curve of petrophysical groups that will support the property distribution in our 3D model.

You can access this article if you purchase or spend a download.