Abstract
Tight sandstone reservoirs characterization and evaluation is very difficult based on conventional well log data owing to the extremely low porosity and permeability, and strong heterogeneity. The main accumulation spaces of conventional reservoirs are intergranular pores, and the pore size is the main controlling factor of permeability. However, besides intergranular pores, fractures play much greater important role in accumulating hydrocarbon, improving the pore connectivity and pore structure in tight sandstone reservoirs. Hence, it should be accurately predicted the pore structure dredged by fractures to improve the characterization of tight sandstone reservoirs. Generally, nuclear magnetic resonance (NMR) logging is an effective method to evaluate formation pore structure. However, it cannot be well used in fractured reservoirs because the NMR T2 spectra has no any response for fractures with width <2mm. The borehole electrical image log is usable in characterizing fractured reservoirs. The pore spectrum, which is extracted from the borehole electrical image log, can be used to qualitatively reflect the pore size. Hence, it will play an important role in fractured reservoirs pore structure characterization. In this study, based on the comprehensive analysis of the pore spectra, the corresponding mercury injection capillary pressure (MICP) data and pore-throat radius distributions acquired from core samples, a relationship that connects the 1/POR and capillary pressure (Pc) is proposed. Established a model based on formation classification to transform porosity spectrum into pseudo capillary pressure curve. In addition, a Swanson parameter-based permeability prediction model is also developed to extract fractured formation permeability. Meanwhile, to verify the superiority and otherness of borehole electrical image and NMR log, the model that evaluated reservoirs pore structure from NMR log is also established. Based on the application of the proposed method and models in actual formations, the evaluated pore structure parameters and permeabilities from two types of well log data are compared. The results illustrates that in formations with relative good pore structure, the predicted pore structure parameters and permeabilities from these two types of well log data agree well with the drill stem testing data and core-derived result. However, in low permeability sandstones with relatively poor pore structure, the porosity spectra can be well used to evaluate the pore structure, whereas the characterized pore structure from NMR log is overestimated. With the comprehensive research of reservoirs pore structure and permeability, the fractured tight sandstone formations with development value are precisely identified. This proposed method has greatest advantages that the pore structure of fractured reservoirs can be characterized, and the contribution of fractures to the pore connectivity and permeability can be quantified. it is usable in tight sandstone reservoirs validity prediction.