A novel de-noising method for NMR echo data obtained from tight sandstone reservoirs

To improve the signal-to-noise ratio (SNR) of nuclear magnetic resonance (NMR) echo data with SNR below 8 obtained from tight sandstone reservoirs and improve the accuracy of the porosity obtained from T_2 inversion, a novel de-noising method is developed to improve the processing of NMR echo data. The algorithm is based on variational mode decomposition wavelet threshold (VMD-WT). It was found that as the decomposition mode number increases, the center frequency difference between the first mode and the second mode gradually decreases. A method to determine a reasonable mode decomposition number based on the center frequency difference of the modes is proposed. The effectiveness of the algorithm is verified by the application in the numerical simulation experiments and actual NMR echo data, respectively. The results indicated that, compared with wavelet threshold de-noising method, VMD-WT de-noising method is more suitable for the de-noising of NMR echo data with SNR below 8 obtained from tight sandstone reservoirs.