Calculating Permeability from Logs

A number of algorithms have been published that estimate permeability from well log data. The model that is recommended is the modified Coates-Denoo technique, which has a variable correlation factor that allows for correlation to real data. It requires inputs of effective porosity and Bulk Volume Water Irreducible (BVWirr). The concept of BVWirr is relatively simple yet very powerful in the estimation of permeability. It is the product of irreducible water saturation and effective porosity, and it represents the volume of the rock that is filled with immobile fluids. It compares to Sw in that Sw is the percent of the effective porosity that is water filled. The nuclear magnetic resonance log is useful in this process because it provides a direct measurement of BVWirr once the T2 cutoff has been calibrated to core data. The calculated permeability from NMR logs should be used with caution, however, because it requires calibration to real data prior to use as a permeability tool. If this calibration has not been done, it is recommended that the modified Coates-Denoo technique be used with only the input of BVWirr from the NMR tool.

The modified Coates-Denoo technique (modified by Barba, 1991): Although the original work used values of 70 and 100 for the C factor, it has been shown in numerous field studies that a variable C factor provides for a better correlation to effective permeability to reservoir fluids (Barba and Darling, 1990; Barba and Batcheller, 2000).