论文部分内容阅读
一种新的数字图像分析方法可以量化从亚微观(普通显微镜下看不到)到毫米级的3个数量级以上范围内的孔隙参数。这种孔隙特征描述方法无需知道样品的岩性、年代、埋深或成岩作用。这种方法是根据光学显微镜(OM)和环境扫描电子显微镜(ESEM)对薄片进行不同倍数的放大而进行数字图像分析。其结果有助于解释各种各样具有复杂孔隙结构的碳酸盐岩样品的渗透率变化。不过,这种分析方法也可运用于其它岩石类型的薄片。光学显微镜(OM)成像提供了大孔隙的信息,而环境扫描电子显微镜提供了微孔隙的信息。大孔隙和微孔隙的界限为500μm~2孔隙面积,转换成孔隙长度约为20μm,这大致相当于薄片的厚度,也就是光学显微镜的分辨率。将数字化的薄片图像二元化为大孔隙和基质相(在OM下)或微孔隙和固相(在ESEM下)。用标准数字图像分析程序来检测全部单孔隙并测量孔隙面积及孔隙周长。基于这些分析就可以计算每个样品的大孔隙数量、基质中的微孔隙(内部微孔隙)量、大孔隙的形状(周长大于面积)及孔隙大小分布。通过比较从岩塞得到的总孔隙度表明,根据这种方法得到的大孔隙度和微孔隙度与岩塞得到的孔隙度相匹配,证明这种方法是可行的。综合大孔隙和微孔隙资料可获取孔隙大小分布和孔隙形状信息,而孔隙大小分布和孔隙形状信息可用来解释岩石物性,尤其是渗透率的分布。通过利用神经网络对参数敏感性的分析表明,在高渗透样品中渗透率似乎主要受大孔隙形状的控制;在低渗透样品中渗透率似乎主要受内部微孔隙数量的控制。
A new digital image analysis method can quantify pore parameters in the range of 3 orders of magnitude above the millimeter from submicroscopic (not visible under normal microscope). This method of pore characterization does not need to know the lithology, age, burial depth or diagenesis of the sample. This method is based on optical microscope (OM) and environmental scanning electron microscopy (ESEM) for different magnifications of the sheet for digital image analysis. The results help to explain the changes in permeability of a wide range of carbonate samples with complex pore structures. However, this method of analysis can also be applied to other rock types of sheets. Optical microscopy (OM) imaging provides macropore information while environmental scanning electron microscopy provides information on microporosity. The boundary between macropores and micropores is 500μm ~ 2 pore area, converted to a pore length of about 20μm, which is roughly equivalent to the thickness of the sheet, that is, the resolution of the optical microscope. The digitized flake images are binarized into macropores and matrix phases (under OM) or microporosity and solid phases (under ESEM). A standard digital image analysis program was used to detect all single porosity and to measure the pore area and pore circumference. Based on these analyzes, the number of macropores per sample, the amount of microporosity (internal microporosity) in the matrix, the shape of macroporosity (circumference greater than area), and pore size distribution can be calculated. Comparing the total porosity obtained from the rock plug shows that the macroporosity and microporosity obtained according to this method match the porosity obtained by the rock plug, proving that this method is feasible. The macroscopic and microporosity data can be used to obtain pore size distribution and pore shape information. The pore size distribution and pore shape information can be used to explain the petrophysical properties, especially the distribution of permeability. Analysis of the parameter sensitivity using neural networks shows that permeability appears to be dominated by macroporosity in high permeability samples; permeability appears to be dominated by the number of internal microporosity in low permeability samples.