PERMEABILITY CONTROLS ON MOLDIC GRAINSTONES: CT BASED PORE NETWORK MODELING OF CARBONATE RESERVOIR ROCK FROM WEST TEXAS AND A NEW NMR-BASED PERMEABILITY TRANSFORM

摘要

We present a study aimed at understanding the main controls on permeability in carbonate grainstones dominated by moldic porosity in an effort to obtain more accurate permeability transforms from NMR logs in such formations. Computed x-ray microtomography (CT) images of the 3-D pore structure of seven core samples from a well in a West Texas oil reservoir revealed that the connectivity of neighboring molds was influenced heavily by flow paths that traverse a microporous matrix of cements in between the molds. A new process to build a pore network model from the CT images was developed. This process involved identification of individual molds and assignment of hydraulic resistivity for each pore-pore connection based on the porosity and thickness of the intervening microporous cement. The resulting models were successful in matching measured permeabilities of the core samples. Subsequent analysis of the pore network model indicated that the average thickness of the walls separating adjacent molds is a good predictor for the permeability. Specifically, the logarithm of the permeability was observed to vary linearly with the average wall thickness. Based on these findings, a new permeability transform was developed to convert NMR T2-distributions to permeability by estimating the average wall thickness from grain size and moldic porosity. The method was validated with 22 further core samples from the same formation, with significantly lower mean-square error than the traditional NMR-based permeability transforms derived for rocks dominated by primary, intergranular porosity.