Numerical Investigations on Nuclear Magnetic Resonance T 1- T 2 Responses in Tight Gas Reservoirs with Oil-Based Mud Filtrate Invasion

Summary The 2D nuclear magnetic resonance (NMR) T1-T2 technique has been widely utilized for fluid identification and reservoir parameter evaluation. However, the invasion of drilling mud filtrate into the reservoir may cause NMR T1-T2 spectra to contain the mud filtrate signal. The reliability of fluid identification and reservoir parameter calculations based on NMR T1-T2 spectra may be reduced. It is therefore necessary to investigate NMR T1-T2 responses for the reservoirs with mud filtrate invasion. In this study, the T1-T2 responses of tight gas rock with oil-based mud (OBM) filtrate invasion were simulated with the random walk method, based on 3D digital core technique. The influences of internal magnetic field gradient, gas hydrogen index (HI), and signal-to-noise ratio (SNR) on T1-T2 responses were also analyzed. The T1-T2 interpretation model for tight gas with OBM filtrate invasion was established, and it was used for fluid identification and saturation calculation in a tight gas reservoir. The results indicate that as the OBM filtrate invasion increases, the OBM filtrate signal gradually strengthens and the gas signal gradually weakens. As the internal magnetic field gradient is enhanced, the water peak, gas peak, and oil peak in T1-T2 spectra shift toward the reduction direction of T2 relaxation time, and the movement amplitude is faster for the gas peak than for the oil peak and water peak. With the gas HI increases, the position of the gas peak in the T1-T2 spectra remains unchanged, while the signal intensity gradually increases. In addition, the reliability and accuracy of the inverted T1-T2 spectra are enhanced with the SNR increases. It is therefore recommended that the SNR of the acquired NMR data should be greater than 20. The established T1-T2 interpretation model is appropriate for the study area, and the calculated gas saturation is consistent with that from conventional logging. These findings may provide a critical theoretical basis for accurate fluid identification and reservoir parameter calculation using NMR logging data with mud filtrate invasion.