Reconstruction of LWD-NMR T2 water spectrum and fluid recognition based on microscopic pore structure constraints

Abstract

The existing fluid property identification methods are not applicable to low resistivity reservoirs. In order to accurately identify low-resistivity oil reservoirs, the effects of fluid properties on the NMR T2 spectrum under different pore structures were analyzed. The results show that the T2 spectrum of the low-resistivity oil section shows a broad and gentle hilly shape, and the T2 spectrum of the water section shows a steep and mountainous shape. Based on the difference characteristics of T2 spectrum of reservoirs with different fluid properties, a new NMR water spectrum reconstruction and fluid property identification method based on microscopic pore structure constraint is innovatively proposed. Based on the drilling NMR logging data, three parameters characterizing the microscopic pore structure are firstly extracted from the NMR T2 spectrum of standard water reservoirs, which reflect the overall shape of the T2 spectrum in terms of transverse relaxation time, longitudinal pore fraction and These three parameters reflect the pore structure, water and oil and gas information of the reservoir in terms of the transverse relaxation time, the longitudinal pore fraction and the overall morphology of the T2 spectrum. Secondly, the microscopic pore structure parameters are used as the coordinate axes to construct a 3D water spectrum sample library, and the standard water formation T2 spectrum samples of the target block are stored in the 3D water spectrum library through the constraints of the three parameters. Finally, the NMR T2 spectrum parameters are extracted from the whole well section and the samples in the water spectrum library are called to realize the water spectrum reconstruction. The parameters that can distinguish oil and water formations are extracted from the measured T2 spectrum and the reconstructed water spectrum and fluid identification is performed. The results show that the reconstructed water spectrum is highly similar to the actual measured T2 spectrum in the production-proven standard water layer, which verifies the accuracy of the water spectrum reconstruction technique; in Caofeidian A oilfield, the fluid identification was performed on 20 production-proven test sections, of which 18 reservoirs were correctly identified, and the identification compliance rate reached 90%. The method circumvents the influence of pore structure in NMR T2 spectrum and highlights the information of its fluid properties, which improves the fluid property identification capability of logging with drilling (LWD) NMR logging.