基于核磁共振测井的致密砂岩储层孔喉参数表征新方法OA

To address the problems of limited prediction accuracy and high inversion uncertainty of low signal-to-noise ratio echo data when traditional nuclear magnetic resonance(NMR)logging T2 spectra are converted into pseudo-capillary pressure curves for characterizing pore-throat parameters of tight sandstone reservoirs,and to provide more accurate reservoir pore-throat parameter support for oil and gas exploration and development,carbon dioxide geological storage,and other applications,the study adopts the time-domain analysis technology of NMR logging.Three groups of Laplace transform function pairs are constructed,and the research is conducted by combining numerical simulation,core experiments,and well logging data processing.The research results are as follows.First,three types of time-domain parameters,namely P1,P2 and P3,are successfully extracted directly from NMR echo data,thus avoiding the T2 spectrum inversion process.Numerical simulation verification indicates that under the signal-to-noise ratios of 30 and 5,the root mean square errors(RMSE)of characteristic parameter calculation by this method are lower than those by the T2 domain analysis method,with more significant advantages under low signal-to-noise conditions.Second,after optimization via core experimental data analysis,the correlation coefficients(R2)between P2 and the average pore-throat radius,median pore-throat radius,and maximum pore-throat radius reach 0.92,0.81,and 0.89,respectively.These values are significantly superior to those corresponding to P1 and P3.Third,for the pore-throat parameter characterization model established based on P2,the R2 for the prediction of the above three key parameters are 0.90,0.76,and 0.82,with the respective RMSE values being 0.03 μm,0.03 μm,and 0.15 μm.The prediction accuracy of this model outperforms that of traditional pseudo-capillary pressure curve conversion methods,including linear conversion,piecewise power function,and extended power function approaches.Finally,processing results of field logging data reveal a high degree of consistency between the predicted pore-throat parameters and the measured results from core experiments.This method thus enables the continuous characterization of pore-throat parameters in tight sandstone reservoirs.It is concluded that the novel method based on time-domain analysis of NMR logging exhibits high stability and excellent anti-noise interference capability.It provides a reliable technical tool for the characterization of pore-throat parameters in unconventional reservoirs such as tight sandstones,and holds important reference value for reservoir quality evaluation of complex reservoirs.