Ghomeshi Shahin, Kryuchkov Sergey, Kantzas Apostolos
University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4, Canada; Perm Inc., VP Technology, 3956 29 Street NE., Calgary, AB T1Y 6B6, Canada.
Perm Inc., VP Technology, 3956 29 Street NE., Calgary, AB T1Y 6B6, Canada.
J Magn Reson. 2018 Apr;289:79-91. doi: 10.1016/j.jmr.2018.02.007. Epub 2018 Feb 14.
Nuclear Magnetic Resonance (NMR) is a powerful technique used to characterize fluids and flow in porous media. The NMR relaxation curves are closely related to pore geometry, and the inversion of the NMR relaxometry data is known to give useful information with regards to pore size distribution (PSD) through the relative amplitudes of the fluids stored in the small and large pores. While this information is crucial, the main challenge for the successful use of the NMR measurements is the proper interpretation of the measured signals. Natural porous media patterns consist of complex pore structures with many interconnected or "coupled" regions, as well as isolated pores. This connectivity along the throats changes the relaxation distribution and in order to properly interpret this data, a thorough understanding of the effects of pore connectivity on the NMR relaxation distribution is warranted. In this paper we address two main points. The first pertains to the fact that there is a discrepancy between the relaxation distribution obtained from experiments, and the ones obtained from solving the mathematical models of diffusion process in the digitized images of the pore space. There are several reasons that may attribute to this such as the lack of a proper incorporation of surface roughness into the model. However, here we are more interested in the effects of pore connectivity and to understand why the typical NMR relaxation distribution obtained from experiments are wider, while the numerical simulations predict that a wider NMR relaxation distribution may indicate poor connectivity. Secondly, by not taking into account the pore coupling effects, from our experience in interpreting the data, we tend to underestimate the pore volume of small pores and overestimate the amplitudes in the large pores. The role of pore coupling becomes even more prominent in rocks with small pore sizes such as for example in shales, clay in sandstones, and in the microstructures of carbonates.
核磁共振（NMR）是一种用于表征多孔介质中流体及其流动的强大技术。NMR弛豫曲线与孔隙几何形状密切相关，已知通过存储在小孔和大孔中的流体的相对振幅，对NMR弛豫测量数据进行反演可得到有关孔径分布（PSD）的有用信息。虽然此信息至关重要，但成功使用NMR测量的主要挑战在于对测量信号的正确解释。天然多孔介质模式由具有许多相互连接或“耦合”区域以及孤立孔隙的复杂孔隙结构组成。沿喉道的这种连通性会改变弛豫分布，为了正确解释这些数据，有必要深入了解孔隙连通性对NMR弛豫分布的影响。在本文中，我们讨论两个要点。第一点涉及这样一个事实，即从实验获得的弛豫分布与从孔隙空间数字化图像中求解扩散过程数学模型获得的弛豫分布之间存在差异。有几个原因可能导致这种情况，例如模型中缺乏对表面粗糙度的适当纳入。然而，在这里我们更感兴趣的是孔隙连通性的影响，并理解为什么从实验获得的典型NMR弛豫分布更宽，而数值模拟预测更宽的NMR弛豫分布可能表明连通性较差。其次，根据我们解释数据的经验，由于没有考虑孔隙耦合效应，我们往往会低估小孔的孔隙体积并高估大孔中的振幅。孔隙耦合的作用在具有小孔径的岩石中变得更加突出，例如在页岩、砂岩中的粘土以及碳酸盐的微观结构中。
