Estimating Klinkenberg-Corrected Permeability From Mercury-Injection Capillary Pressure Data: A New Semianalytical Model for Tight Gas Sands

摘要

This paper presents the practical applications of a semi-analytical model for estimating the Klinkenberg-corrected per-meability from mercury-injection capillary pressure (Hg-pc) data in tight gas sands. The fundamental relationships be-tween rock pore size/geometry and basic rock properties are well-documented in the petroleum literature. Moreover, since rock pore characteristics can be accurately quantified from interpretation of mercury-injection capillary pressure data, the literature is replete with models for estimating permeability from Hg-pc data. However, existing Hg-pc models tend to yield inconsistent results — and few models have been shown to be directly applicable for low-permeability sands. The basis of our model is the Purcell/Burdine model (bundle of capillary tubes) combined with the Brooks/Corey model (power law relationship of capillary pressure versus wetting phase saturation). We tested our model using more than 100 sets of mercury-injection capillary pressure data. Effective porosity in our data set ranges from 4 to 14 percent, while absolute permeability ranges from 0.005 to 0.5 md. The primary technical contribution of this paper is a tuned model for estimating Klinkenberg-corrected permeability from mercury-injection capillary pressure (Hg-pc) data in tight gas sands. The final form of our model allows estimation of the absolute (Klinkenberg-corrected) permeability as a function of effective porosity, irreducible wetting phase saturation, dis-placement pressure, and pore size characteristics. The model is also reversible — we can estimate a capillary pressure profile from routine permeability and porosity data.