An improved method in petrophysical rock typing based on mercury-injection capillary pressure data

Abstract

ABSTRACT Petrophysical rock typing plays a major role in carbonate reservoir characterization and geological modeling for reservoir simulation. In this study, based on 1120 samples of mercury-injection capillary pressure data, we have analyzed the contributions to rock permeability for the thick Cretaceous carbonate reservoir of Mishrif Formation in H Oilfield in Iraq. The analyses have been carried out for three types of pore-throat frequency distribution, i.e., monomodal, bimodal, and trimodal. The results show that for the bimodal and trimodal samples, the pore-throat frequency distribution of the first pore system contributes more than 95% of rock permeability. Further, the Thomeer function has been used to fit the mercury-injection capillary pressure curve, in which the displacement pressure, pore geometry factor, and bulk volume have been extracted. These three parameters have been utilized to characterize the pore structure and to fit with the rock permeability so as to obtain the reservoir classification parameter Mode. Then, we have established the relationship between permeability and mode, permeability and R35, and the correlation coefficients are 0.81 and 0.75, respectively. And permeability derived from Mode is more accurate than R35. The Modes of 1120 samples have been arranged in ascending order and plotted on the semi-logarithmic scales. After interpolation, derivation, and filtering, five rock types have been divided at the change in slope. The combination of Mode and this slope method is considered an improved method. The classification results of Mode, Winland R35, and Flow Zone Indicator have been compared with those of pore-throat distribution, T2 distribution of nuclear magnetic resonance, and conventional logging responses. The results demonstrate that the classification results of Mode are significantly better than those of Winland R35 and Flow Zone Indicator.