Integration of Computed Tomography Scanning and Nuclear Magnetic Resonance Results with Conventional Laboratory Test Data for Effective Reservoir Characterization

Abstract

Abstract The paper suggests a new laboratory tests workflow, aimed at reservoir characterization and rock typing of a polimineral clastic formation, based on microstructural characteristics. The proposed approach is accompanied by a comparative study of various laboratory techniques, including CT, NMR, MICP, centrifuging, and thin section analysis. The methodology for combining quantitative porous medium characteristics, derived from these multi-scale tests is also provided. The developed laboratory test program includes porosity and permeability measurements, capillary pressure, electrical resistivity, NMR and CT-scanning tests. For initial differentiation of samples based on porosity, permeability and irreducible water saturation, we applied methods of flow zone indicator (FZI) and Winland R35. Afterwards, we performed joint interpretation of FZI with the results of centrifuging, resistivity, NMR and CT-scanning tests. Finally, for the most representative samples from the rock types, defined at previous steps, we performed X-ray fluorescence spectroscopy, XRD, MICP and thin section analyses. In our study, we applied FZI, calculated as a function of irreducible water and porosity. Compared to conventional FZI, derived from porosity and permeability, our approach enabled more reliable rock typing in both sandstones and low-permeable siltstones. CT tests showed that within the distinguished 4 rock types grain and pore sizes increase from the 1st to the 4th rock types. CT-scanning and X-ray fluorescence spectroscopy detected presence of pyrite in all samples. Siltstones contain more pyrite than sandstones that reflects deeper marine and consequently more anoxic depositional environment. The obtained NMR T2-spectra appeared to be effected by both porous medium and mineralogical composition. Pyrite inclusions caused shifting of the T2-spectra to lower values, while carbonate presence – to higher ones. Therefore, since NMR signal is effected by multiple factors, applying of a single T2cutoff value for reservoir characterization and rock typing can lead to ambiguous interpretation. The proposed algorithm for combining quantitative parameters of the porous medium allowed integration of pore size distributions, derived from CT and capillary tests. This enabled characterization of the whole range of pore sizes covered only partly by the different methods. The resulted pore size distribution enabled calibrating NMR Т2 spectra and defining individual surface relaxivity values for each rock type. The selected set of techniques allowed confident differentiation and characterization of the rock samples based on multiple parameters, which were interpreted in relationship with each other. Moreover, the developed methodology was verified by applying additional laboratory tests, which confirmed the results of the proposed rock typing and reservoir characterization.