New Method for Nuclear Magnetic Resonance Dual Echo Time Logging Interpretation and Fluid Identification in Tight Sandstone Gas Reservoirs

Summary For tight sandstone gas reservoirs, conventional logging has great uncertainty, and fluid identification is difficult. Nuclear magnetic resonance (NMR) dual echo time (TE) logging has good application potential in identifying tight sandstone gas layers. The original NMR logging spectrum and diffusion analysis (DIFAN) methods both assume that the diffusion coefficients of gas and water are known and reflect gas information. However, the existing methods cannot be applied to complex gas layers. In this paper, we propose a new NMR dual TE logging interpretation method combining particle swarm optimization (PSO) with least squares (LSQR) algorithm inversion. Based on the principle of NMR dual TE logging activation, we analyze its relaxation mechanism and derive the equations between the echo train amplitude difference and fluid type, fluid content, and fluid diffusion coefficient. Based on the PSO inversion method, we perform nonlinear inversion of the constructed objective function to obtain parameters, such as the diffusion coefficient of the fluid, and generate an echo train difference that reflects the fluid information. We then use the LSQR method to invert the generated echo train differences to obtain a difference spectrum that reflects hydrocarbon information, and then calculate the gas volume and gas saturation of the reservoir’s flushed zone. We verify the correctness of the new method by building a rock model and conducting numerical simulations. Based on the calculation results, the gas-containing porosity in the flushed zone calculated by the new method and the gas-containing porosity in the flushed zone calculated by the time domain analysis (TDA) method crossplot are constructed. We applied the above method to the Huagang formation in a basin in eastern China. The results show that this idea and the inversion method can effectively realize the NMR dual TE logging data processing of tight sandstone gas reservoirs, and the fluid identification crossplot established based on the processing results improves the accuracy of gas layer identification.