Study on 3D Micro-pore Structure Characteristics of Dolomite Reservoir with Different Scales

an-Abstract Tight dolomite reservoir microscopic pore structure characteristic is the key factor to explore this kind of unconventional reservoirs. In this paper, 3D high resolution reservoir characterization technique based on Micro - CT imaging, constant-rate mercury injection, high-pressure mercury injection and low-temperature nitrogen adsorption test technology are applied to study the micro-pore textures characteristics for T12C well Tanggu, Dagang oilfield with different scales. Through Micro-CT scan imaging, image processing, 3D visualization, porosity calculation and pore structure parameter reconstruction, it is concluded that Micro - CT imaging can directly reflect the changes of fracture and pore in 3D space; the pore radius distribution are close to the result of constant-rate mercury injection and high-pressure mercury injection method, wider than that of low-temperature nitrogen adsorption test; compared with these three methods, the porosity range obtained by Micro - CT imaging technology is slightly narrow. Thus, in order to be effective, comprehensive and accurate to characterize different scales microscopic pore structure characteristics of tight dolomite reservoir, it is better to integrate these four methods.


Introduction
S 3 5 interval of Tanggu oilfield, located in Dagang, Tianjin, is a tight oil reservoir, with the character of abnormal pressure, low porosity and permeability. Dolomite, argillaceous dolomite is dominant within the reservoirs; reservoir space is the rock cracks and pores; low natural productivity, taking advantage of fracturing can improve productivity. The character of this kind of special lithologic body is poor reservoir property, small pore throat radius, the high displacement pressure, which seriously restricts fluid flow and development, so new ways and technologies should be applied in the fine reservoir description, to determine the micro pore structure of reservoir, finally support the oilfield development. Using the Micro -CT technology to carry on 3D scan and reconstruction for 5 cores in T12C well, combined with constant-rate mercury injection, high-pressure mercury injection and low-temperature nitrogen adsorption test technology, the paper analysis the microscopic pore structure of dolomite reservoir in Tanggu oilfield.

Sample Design
The samples are derived from dolomite and argillaceous dolomite reservoir cores of S 3 5 interval of Tanggu oil field in T12C Dagang oilfield, before CT scanning and microscopic pore structure analysis, first using the microscope to identify the rock section and determine porosity, permeability. The identification of thin section shows that the samples are consist of ankerite, feldspar, zeolite and pyrite, with some sections developing fracture, solution pores, and intergranular hole, and other region's pore poor. Use the AP -608 automatic pressure analysis of porosity and permeability tester produced by Coretest company in America to analysis, the porosity of the samples is 12.3%, and the permeability is 2.24×10 -3 μm 2 . The sample is processed into a 2-7mm diameter subsamples, then carry on high resolution Micro-CT scan, the corresponding pixel size is 4.95 mm, maximum resolution of 0.5 mm.

CT Imaging
In order to ensure that 3D pore structure analysis result is consistent with the actual pore structure, image resolution should be as high as possible. According to the need of this research, the instrument for CT scanning the sample of dolomite core is the type 1172 Micro-CT, produced by Belgian Skyscan company, with X-ray source voltage 20～80KV, electric current 0～100μA, diameter of focus≤8μm, the power of tube ≥8W, the highest resolution ≤0.5μm/5μm (10% MFT, isotropic). X-ray detector can obtain CT image of 4000×2600 pixels, the depth of each pixel is 16. The an-gular accuracy of the sample turn table is 0.001°; the Pixel resolution of dolomite CT image is 5μm (Figure 1). The advantage of the instrument is that which can reconstruct 2D and 3D images and be saved in the form of animation and photo, define any shape of the interest area, support 3D image segmentation and stitching. Observed from (Figure 1), the pore of sample M1 to M2 develops differently. The facture of sample M1 develops well, with big pore in some regions; the sample M2 is tight, have strong heterogeneity, with big pore in some regions; the sample M3 contains flower mineral, develops dissolved pore; the lithology of the sample M4 is dense, with good development of mineral and strong heterogeneity; the sample M5 is homogeneous, poor mineral.   According to the scanning image, take advantage of VG Studio MAX2.0 software to render solid phase ( Figure 2) and pore phase 3D visualization image ( Figure 3) of dolomite subsample. As depicts in Figure 2, the sample obtained from T12C well have good quality of fracture, with dissolved pore and inter crystalline pore as the main reservoir space, narrow throat between pores, small coordination number. Through the software, any slice can be made into 3D visualization image, then display pore space in any direction, thus we can know the sample's fracture morphology diversity, the sizes. Mineral particles in two of five samples, clear, in form of preserved egg's surface, are analcime after microscopic identification and elemental chemical analysis. Dissolution pore developed well within the mineral, and later was filled by outside material. Compared with injection experiment and transmission electron microscopy, Scanning Electron Microscopy (SEM) and other traditional experimental analysis method, through the establishment of three-dimensional visualization images of the rock pore structure, we can be more intuitive to study pore space dis-tribution of reservoir rock, especially the pore connectivity, then use statistics of pore and throat, further study the distribution and connectivity of microscopic pore in the three dimensional space within the core. Take advantage of analysis software CTAn to study the parameters of 5 samples which have already been scanned by Micro-CT, pore diameter as shown in Table 1. After binarization processing, the software can also calculate closed porosity, open porosity and total porosity of the image and average pore diameter. As we can see from Table 1, the porosity of 5 samples is between 2.39 and 6.31%, the average porosity is 4.112%. In a single image, minimum porosity was 1.66%, the porosity is 18.26%; the aperture size of 5 samples is between 7.37 and 10.8μm, with average pore diameter 8.726μm, the minimum aperture 0.96 μm, the maximum diameter 33.8μm ( Table 1). The porosity is 2.39-6.31%, calculated by the way of Micro-CT, with the average porosity of 4.112%, significantly less than pressure porosity (12.3%) conventionally measured. Because it just calculates matrix porosity of the imag-es scanned by Micro-CT, regardless of the cracks and microcracks, meanwhile, there is no obvious relationship between porosity and aperture. It is consistent with the conclusion proposed by Jingbohefu [1] who also considered it no correlation between porosity and dolomite pore aperture [2]. Figure 4 shows the frequency of the main pore diameter of T12C well, the pore diameter distribution concentrates among 6-23.9μm, also existing above 23.9μm, belongs to small-medium aperture scope, with small coordination number. It indicates that the submicron level, micro pore and throat have not been distinguished by Micro-CT. Therefore, when we apply the related statistical result of the pore structure based on Micro-CT images, but also refer to the mercury injection method, low temperature nitrogen isothermal adsorption method, and other test data to verify [3][4][5][6][7].     Constant-rate mercury injection is that meniscus of mercury front will generate nature pressure fluctuation when it flows across micro-pore of different shapes by means of constant-rate mercury injection, in the process of quasi static, the pore microstructure is determined. Using ASPE-730 TM (Automated System for Pore Examination) constant-rate mercury injection apparatus to test 5 samples, and their micro homogeneous coefficient, average throat radius, and the contribution of single throat to permeability, relative sorting coefficient and other important characteristic parameters are obtained by ASPEDAS analysis software applying as well [8][9][10]. The analysis results are shown in Table  2 and Figures 5-7. As we can see from Table 2, the average throat radius of 5 samples is 3.563μm, which is close to 8.726μm calculated by Micro-CT. As it shown in Figure 5, the curve is in form of jagged pattern; the throat radius less than 2μm is slightly predominant; the smaller the permeability is , the closer the peak is to the throat area, the distribution frequency becomes smaller from throat to big throat, but the trend is not obvious. The large aperture is mainly concentrated on 80-160μm, certainly some pore radius existing in the scope of 160～400μm, consistent with the Micro-CT analysis results(the radius belonging to the range of 79.3 ～158.5μm takes up 10.1%, Table 1, Figure 1); the difference of radius throat' effect on permeability is big, as Figure 7. Jagged distribution curves impress that the range of the throat distribution of T12C is large, no obvious relationship with permeability. As it shown in constant-rate mercury injection experiment, the pore radius distribution of different permeability core is similar (Figure 6), however, the radius of throat varies widely ( Figure 5). It suggests that the throat distribution is the key factor to determine the quality of permeation, if the reservoir permeability is up to the large throat, then the larger the seepage channel, the flow resistance is smaller, the stronger the seepage ability is, the greater the development potential of reservoir; on the contrary, the reservoir permeability is determined by small throat, then the fluid seepage resistance is big, the seepage ability is weak, correspondingly the development of reservoir very difficult ( Figure 7).

Study the Pore Structure based on High Pressure Mercury Intrusion Method
During the experiment of high pressure mercury intrusion, it will form the pressure curve of the intrusion mercury volume to pressure and the withdraw mercury volume to pressure, the curve of the pore radius to the volume distribution density, the curve of the pore radius to normalized volume distribution curve. High pressure mercury intrusion method is that study the reservoir micro pore structure with the pore throat radius and distribution and other micro pore structure information provided by these curves [11][12][13][14][15]. 5 samples of the T12C well were selected to explore the structure of the pore. The instrument is Autopore IV 9500 high pressure mercury injection apparatus imported from America, the pore diameters distribution ranges from 360μm to 0.005μm, the test pressure range is 2KPa～228 MPa, and the test scope covers the micro-pore , meso pore and macro pore(in accordance with the pore classification standard of the international federation of pure chemistry and applied chemistry (IUPAC), 0 to 2nm is micro-pore; 2-50 nanometer is meso pore, more than 50nm is macro pore). Table 3 show the test results of 5 samples, Figure 8 displays the capillary pressure curve of GM4 sample. As it shown in the Figure 8, the mean pore radius is 0.04μm，based on the classification standards of IUPAC, GM4 sample belongs to the meso level. From Table 3, the reservoir pore radius distribution ranges from 0.034μm to 3.91μm; intrusion Hg saturation and extrusion Hg efficiency have no obvious relationship with porosity and permeability, but evident with the mean radius and lithology. Nanoscale and sub-micron pore throat of dolomite rock sample is much better than that of argillaceous dolomite rock sample argillaceous dolomite pore structure develops better than that of dolomite [16][17][18][19][20].  Low-temperature nitrogen adsorption method is based on physical adsorption principle of nitrogen in solid surface, measured nitrogen adsorption capacity under different relative pressure, drawing the sorption isotherm. It is through BET (Brunauer Emmett Teller) equation or langmiur to calculate monolayer adsorption capacity, to acquire the specific surface area of the sample, then makes use of BJH (Barrot -Joyner Halenda) method to calculate the pore size distribution. This method can detect the aperture distribution range which is 0.00035～0.1μm nanoscale pore throat, test range of microporous and mesoporous level, the highest accuracy less than 0.05μm. Table 4 and Figure 9 show the 5 sample of T12C well low-temperature nitrogen adsorption test result. Pore volume in Table 4 is that the percentage of micro-pore and mesopore volume takes up unit mass rock sample, distribution range of 16.57 to 31.84 mm 3 /g, pore volume on average value of 24.22 mm 3 /g. In the view of lithology, dolomite and argillaceous dolomite is slightly different: the pore volume of dolomite is between 20.93 and 31.84 mm 3 /g, with a mean of 26.38 mm 3 /g; pore volume of argillaceous dolomite range from 16.57 to 31.55 mm 3 /g, with a mean of 22.78 mm 3 /g, slightly less than the pore volume of argillaceous dolomite. It illustrates that dolomite pore space under 0.05μm is larger than that of argillaceous dolomite, with fluid flowing difficultly. The average radius of micro-pore and mesopore in Table 4 ranges 0.0071 to 0.027μm, dolomite's on average is smaller.

The Low-Temperature Nitrogen Adsorption Method to Study the Pore Structure
The results are consistent with the conclusion drawn by the high-pressure mercury intrusion method and illustrate that Micro -CT cannot distinguish the micro pore throat of the submicron level inside the core. The porosity of the meso-microporopus is the ratio of meso-microporopus volume to the volume of the core. The porosity of the meso-microporopus of 5 samples in T12C well ranges from 3.70% to 7.26%, with the average value of 5.43%, which illustrates that the porosity of T21C well is low, and the conclusion is highly consistent with the porosity of 2.39-6.31% and the mean value of 4.112% measured by the way of Micro-CT. The volume ratio of the meso-microporopus, another parameter in Table 4，is defined as the ratio of volume of pore, with its radius less than 0.05μm, to the total volume of the pore. The volume ratio of the meso-microporopus of the five samples is 27.73-55.0%, and the average value is 42.68%, which implies that about half reservoir space scatters in the pore throat less than 0.05μm, resulting in the development difficult. The volume ratio of the meso-microporopus of dolomite ranges 43.85% to 57.30%, with the average value of 50.58%; the argillaceous dolomite's is 27.73-55.0%, and the mean of 37.42%. Apparently, most of the spatial of dolomite distributes in the pore and throat less than 0.05μm, and significantly greater than argillaceous dolomite, therefore, argillaceous dolomite has better development potential than dolomite [21,22].
The S-V, in Table 4, is the ratio of the total surface area to the volume of core sample. The size of S-V is related to the intensity of solid and liquid effect: the greater the S-V is, the more intense the effect of solid and liquid effect is, and the weaker of the capacity of low, more difficult to develop. The S-V of the 5 samples is 0.69-1.97 (×10 7 /m -1 ) with the average value of 1.04 (×10 7 /m -1 ). Among them, the S-V of dolomite is in the range of 0.74 to 1.97 (×10 7 /m -1 ), with the mean of 1.36 (×10 7 /m -1 ); the argillaceous dolomite's ranges from 0.69 (×10 7 /m -1 ) to 0.94 (×10 7 /m -1 ) ,with the mean of 0.82 (×10 7 /m -1 ).Thus, the solid-liquid effect of dolomite is stronger than argillaceous dolomite, and development difficulty is relatively large [23][24][25]. As it shown in Figure 9, specific surface test curve of T12C well can be divided into four types (Ⅰ, Ⅱ, Ⅲ, Ⅳ) as following: Type Ⅰ: The overall state of multi-peak, vertex occurs at the microporous and mesoporous border area, mesoporous area shows a small fluctuation and rises to the similar height with the peak of meso-microporopus (D23-1).
Type Ⅱ: Microporous interval appears small-scope fluctuation, with local pronounced peaks, as a whole, present the bimodal or many peaks, but the vertices is in the range of mesoporous (D32-1).
Type Ⅲ : Microporous interval appears small-scope fluctuation, with local pronounced peaks, overall present the bimodal or many peaks, but the vertices in the range of mesoporous (DN3-2).
Type Ⅳ: The pores at all levels, gradually increase from small to large, with no obvious jump (DM5-3).

Conclusion
a) Based on all the above research, it can be concluded as followings: b) Micro-CT scanning is a good tool to describe the micro-network of pore structure. Pore space, pore throat radius, closed porosity, open porosity and total porosity can be calculated; also coordinate number is estimated as well. c) Compared with Constant speed mercury intrusion method, High pressure mercury intrusion method and low-temperature nitrogen adsorption method, Micro-CT scanning is most adequately to the T12C well to study the micro-network of pore structure, especially for theμm size pore scale.
d) The pore space radium distribution of the T12C well range from 6 to 23.9μm, moreover more than 23.9μm is well. The mean porosity is 4.1% that belongs to small scale type of porosity. e) As a result of constant speed mercury intrusion method, high pressure mercury intrusion method and low-temperature nitrogen adsorption method, argillaceous dolomite has better development potential than that of the dolomite.