期刊
JOURNAL OF MAGNETIC RESONANCE
卷 269, 期 -, 页码 196-202出版社
ACADEMIC PRESS INC ELSEVIER SCIENCE
DOI: 10.1016/j.jmr.2016.06.013
关键词
Diffusion; Pore size distributions; Porous media; DDIF; Surface relaxivity
资金
- Science Foundation of China University of Petroleum, Beijing [2462015YJRC003]
- National Natural Science Foundation of China [41130417, 21427812]
- 111 Program [B13010]
The time-dependent diffusion coefficient (D) is a powerful tool to probe microstructure in porous media, and can be obtained by the NMR method. In a real porous sample, molecular diffusion is very complex. Here we present a new method which directly measures the relationship between effective diffusion coefficients and pore size distributions without knowing surface relaxivity. This method is used to extract structural information and explore the relationship between D and a in porous media having broad pore size distributions. The diffusion information is encoded by the Pulsed Field Gradient (PFG) method and the pore size distributions are acquired by the Decay due to Diffusion in the Internal Field (DDIF) method. Two model samples were measured to verify this method. Restricted diffusion was analyzed, and shows that most fluid molecules experience pore wall. The D(a) curves obtained from correlation maps were fitted to the Pade approximant equation and a good agreement was found between the fitting lines and the measured data. Then a sandstone sample with unknown structure was measured. The state of confined fluids was analyzed and structural information, such as pore size distributions, were extracted. The D - T-1 correlation maps were also obtained using the same method, which yielded surface relaxivities for different samples. All the experiments were conducted on 2 MHz NMR equipment to obtain accurate diffusion information, where internal gradients can be neglected. This method is expected to have useful applications in the oil industry, particularly for NMR logging in the future. (C) 2016 Elsevier Inc. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据