It is difficult to calculate the bound water saturation and water saturation directly by using Archie method in the low contrast of oil and water layers in marine reservoirs. In order to solve the problem,based on geologic factors,core and logging data,combined with logging response characteristics of low resistivity reservoir,the formation mechanism of marine low resistivity reservoir was studied,and the fine evaluation model of reservoir saturation was established by analyzing the main controlling factors. The results show that the bound fluid and additional conductivity characteristics of glauconite are the main cause of the formation of low resistivity reservoir. The low resistivity reservoir saturation calculated by bound water saturation calculation model based on the main controlling factors and the corrected water saturation calculation model based on conductive factor are consistent with bound water saturation analyzed by core mercury injection experiment. Applied to marine low resistivity reservoir evaluation in Oriente Basin,the corrected water saturation calculation model based on conductivity factor improved the coincidence rate of interpretation and reservoir parameter interpretation accuracy,and achieved good application effect, verifying its effectiveness.

以四川盆地西南部二叠系峨眉山玄武岩组基性火山岩为例，基于基性火山岩的岩石学特征及核磁共振弛豫机理，通过内部磁场梯度数值模拟方法及变回波间隔核磁共振实验，分别探讨了内部磁场梯度、顺磁性矿物含量对T₂谱的影响，提出了一种表征基性火山岩孔隙结构的新方法。研究结果表明：①四川盆地西南部二叠系峨眉山火山岩储层岩性主要为玄武质火山碎屑熔岩、灰质角砾熔岩和玄武岩，灰质角砾熔岩和玄武质火山碎屑熔岩的矿物成分均主要为方解石、石英、斜长石及黏土矿物，黏土矿物的平均质量分数分别为27%和32%，其中绿泥石在黏土矿物中的占比分别为84%和33%。②基性火山岩中顺磁性矿物（绿泥石+含铁矿物）含量较高，在核磁共振测量时会产生较强的内部磁场梯度，在高内部磁场梯度下，随着回波间隔的增大，T₂谱主峰向短弛豫位置移动，谱面积不断减小，核磁孔隙度偏小；内部磁场梯度值越大，孔隙的几何形态越扭曲，孔径越小；内部磁场梯度对灰质角砾熔岩影响最大，其次为玄武质火山碎屑熔岩，对辉绿玢岩的影响最小。③通过数据拟合建立基于核磁孔隙度相对误差与顺磁性矿物含量的孔隙度校正公式；利用纵向弛豫时间T₁几乎不受内部磁场梯度影响的特点，将T₁转化为孔径分布；建立T₁与T₂谱几何平均值的关系，对T₂谱的峰值移动幅度进行校正，再进行孔径分布转换，即可实现T₂谱核磁孔隙结构评价。④通过该方法计算的研究区核磁孔隙度与测井孔隙度的相对误差为15%，平均孔喉半径与CT数字岩心实验得出的平均孔喉半径的误差为6%，研究区火山岩孔隙分布非均质性强，以中小孔喉为主。

Taking Permian Emeishan Basalt Formation in southwestern Sichuan Basin as a case study，based on petrological characteristics and nuclear magnetic resonance relaxation mechanism of basic volcanic rocks，the effects of internal magnetic field gradients and paramagnetic mineral content on <i>T</i>₂ spectra were analyzed by using a combination of numerical simulations of internal magnetic field gradients and variable echo time NMR experiments. A new method for characterizing the pore structure of basic volcanic rocks was proposed. The results show that：（1）The volcanic reservoirs in the study area are mainly composed of basalt volcanic clastic lava，calcareous breccia lava，and basaltic rock. The mineral composition of both calcareous breccia lava and basaltic volcanic debris lava is mainly composed of calcite，quartz，plagioclase，and clay minerals. The average mass fractions of clay minerals are 27% and 32%，respectively，with chlorite accounting for 84% and 33% of clay minerals，respectively.（2）The content of paramagnetic minerals（chlorite+iron containing minerals）in basic volcanic rocks is relatively high，which can generate strong internal magnetic field gradients during NMR measurements. Under high internal magnetic field gradient，as echo time increases，the main peak of the <i>T</i>₂ spectrum shifts toward shorter relaxation times，the overall spectrum area gradually decreases，and lead to a smaller nuclear magnetic porosity. The larger the internal magnetic field gradient value，the more distorted the geometric shape of the pores and the smaller the pore size. The internal magnetic field gradient has the greatest impact on calcareous breccia lava，followed by basaltic volcanic debris lava，and has the least impact on diabase porphyry.（3）Establishing a porosity correction formula based on relative error of nuclear magnetic porosity and paramagnetic mineral content through data fitting，converting <i>T</i>₁ spectrum into pore size distribution based on the feature that <i>T</i>₁ measurement is almost not affected by internal gradient magnetic field，establishing the relationship between the geometric mean value of <i>T</i>₁ and <i>T</i>₂ spectrum，correcting the peak shift of the <i>T</i>₂ spectrum，converting <i>T</i>₂ spectrum into pore size distribution，and then <i>T</i>₂ NMR pore structure evaluation can be achieved（. 4）The relative error between nuclear magnetic porosity calculated by this method and the logging porosity is 15%，the relative error of the average pore throat radius between calculated by the method described and from CT digital core experiment is 6%，the distribution of the basic volcanic rocks in the study area are highly heterogeneous，mainly consisting of small and medium-sized pores throats.