Pore Structure and Fractal Characteristics of Deep Shale: A Case Study from Permian Shanxi Formation Shale, from the Ordos Basin.

Pore structure has certain significance for the preservation and enrichment of shale gas. However, less attention is paid to deep shale (>3000 m) which has unique pore characteristics that distinguish it from the shallow and medium layers. In order to study the pore structure characteristics of deep shale, 10 samples of the Shanxi Formation are collected from well YP-1 within the depth of 3550-3610 m in the Fuxian block of the Ordos Basin. The pore structure characteristics of shale samples are quantitatively studied by scanning electron microscopy (SEM), low-temperature nitrogen adsorption-desorption, and high-pressure mercury injection experiments. The pore surface area (SA) and pore volume (PV) of the deep shale of Shanxi formation are low, with average values of 4.282 m2/g and 0.0126 mL/g, respectively. The content of total organic carbon (TOC) is high, which is in the high over mature stage, with undeveloped organic pores and developed microfractures. The main mineral components are clay (51.6%∼89.1%) and quartz (8%∼41.7%). By establishing the relationship between SA, PV, and TOC for quartz and clay minerals, it is found that these three parameters have little contribution to SA and PV. The pore diameter is mainly mesoporous, 2.5-4 nm and 8-11 nm. The complexity of pore structure is discussed through the fractal dimension calculated by the fractal Frenkel-Halsey-Hill (FHH) model. The pore fractal dimension D 2 (2.6240) is greater than D 1 (2.5608), and the complexity of the pore structure is greater than that of the pore surface. The fractal dimension of deep shale in Shanxi is negatively correlated with TOC content and weakly correlated with quartz and clay minerals. It shows that the mineral composition of deep shale in Shanxi Formation in the study area has little effect on pore development, and the development of microfractures is the main contribution of SA and PV.

Introduction

From the concept of shale gas proposed by the United States in the last century to the introduction of shale gas in China at the beginning of this century, shale gas theory,[1−4] exploration, and development have made many gratifying achievements in China.[5−7] According to previous statistics, the recoverable resources of marine–continent transitional shale gas in China are 8.97 × 1012 m3, and the Ordos Basin accounts for more than half.[8−10] Many sets of shale are developed in Paleozoic and Mesozoic in this basin, among which Yanchang chang 7 shale has obtained industrial gas flow, while Benxi Formation, Taiyuan Formation, and Shanxi Formation have had little breakthrough.[11−15]

The accumulation mechanism of shale gas is complex, and the occurrence modes are various. At present, the production mainly comes from the medium and shallow layers. Some studies predict that the recoverable resources of shale gas with a buried depth of 3000–6000 m in China are 20.93 × 1012 m3,[16] while the deep (3500–4500 m) shale gas is widely distributed, which is important for shale gas exploration and development.[17−19] However, deep shale gas has the characteristics of high on-site gas content, low trial gas production, fast decline, and fast pressure drop.[20−22] Some scholars have pointed out that deep shale has more developed pores and fractures than shallow shale, so the gas-bearing property is better.[23] At present, the research on shale pores mainly focuses on the shallow and medium layers, and the pore structure characteristics of deep shale are not clear.[24,25] Therefore, it is particularly important to study the pore structure characteristics of the shale gas reservoir.

In recent decades, the characterization of shale pore type and structure has changed from the qualitative research stage to the quantitative research stage. Field emission scanning electron microscopy (FE-SEM), focused ion beam scanning electron microscopy (FIB-SEM), wide ion beam grinding, scanning electron microscopy (BIB-SEM), transmission electron microscopy (TEM), and micro- and nano-CT were used for qualitative and semiquantitative analysis of pore structure.[26−29] The pore structure was quantitatively analyzed by the low-temperature N2/CO2 adsorption–desorption method, the mercury intrusion method, nuclear magnetic resonance, and X-ray scattering.[26,30−33] At present, pore structure parameters obtained by low-temperature nitrogen adsorption–desorption and mercury intrusion have been widely used in the FHH model to characterize the complexity of the pore structure of shale with different scales.[34−37] The fractal model (FHH) is used to quantitatively describe the irregularity of the complex system.[38] Because of the complexity of the pore structure, it has been widely used in geopetrology. Fractal theory has been proved to be effective in quantifying the heterogeneity of the pore surface and the complexity of the internal structure.[39,40]

Previous studies on the pore structure of shales by these methods mostly stay in the middle and shallow layers.[31,34−37,41] With the increase of the buried depth of the shale gas target layer, the pore structure characteristics of deep shale need to be studied. In the present study, low-temperature N2 adsorption–desorption, high-pressure mercury injection, and scanning electron microscopy, combined with the fractal Frenkel–Halsey–Hill (FHH) model, are used to quantitatively characterize the complexity of the pore structure of deep shales in the Shanxi Formation. The research in this paper can provide a theoretical understanding for the exploration and development of deep shale gas.

Geological Setting and Samples

The Ordos Basin (OB), one of the largest composite and hydrocarbon-bearing basins, is located in the western part of the North China Plate. Because the OB contains a large number of coal, oil, natural gas, and coal-bed methane reserves, it is considered to be one of the most important fossil fuel energy provinces in central and western China.[42−44] Tectonically, it is a huge asymmetric fold, which can be divided into six secondary structural units: the Yimeng uplift, Weibei uplift, Tianhuan depression, Western edge thrust belt, Jinxi fold belt, and Yishan slope.[45−47] Among them, the Yishan slope covers a large area, which is the main location for oil and gas exploitation in the OB. The study area is located in the south of the Yishan slope (Figure a).

Figure 1

Structural features of the Ordos Basin and the location of the sampling wells (a). Stratigraphic column of the Permian Shanxi Formation in the study area with sampling (b).

Generally speaking, the Upper Paleozoic in the OB was deposited in the marine–continental transitional environment, while the Lower Permian Shanxi Formation was deposited in the transitional stage from the transitional environment to the continental environment.[48] A stratigraphic column of the Permian Shanxi Formation in the study area with sampling is shown in Figure b. Shale samples were collected from the YP-1 well at the depth between 3550 and 3610 m. As can be seen from Figure b, the transitional shales in the study area have a main rock type of mudstone with two layers of coal.

A depth of greater than 3000 m is the critical value for organic matter to mature and enter the hydrocarbon generation stage.[49−51] Shale starts hydrocarbon generation when the reservoir physical properties begin to improve. The rich organic shale began to gradually change from the physical action of mechanical compaction to chemical diagenesis dominated by the transformation of mineral composition. The primary pores dominated by intergranular pores changed to secondary pores. The porosity and permeability of shale began to turn from decreasing with the increase of depth to gradually increasing.[16] With the increase of depth, temperature, and pressure, kerogen and its products gradually produce more natural gas, which promotes the hydrocarbon generation transformation of organic matter and the formation of various secondary pores, which provide a place for shale gas enrichment. The weathering oxidation gradually weakens, which has little impact on the preservation conditions of shale gas. In the Paleozoic shale development area dominated by marine facies or sea land transitional facies, the sedimentary facies, rock mineral composition, and organic carbon content of deep shale are diverse, and the organic matter and composition characteristics of deep and shallow shale are not obvious.[16]

Experimental Methods

Ten samples were collected from the YP-1 well core with a depth of 3550–3610 m. These samples are used to investigate mineralogical, geochemical, and pore structure characteristics. A series of experiments were carried out, including X-ray diffraction (XRD) analysis, low-pressure N2 adsorption–desorption analysis (LP-N2-GA), high-pressure mercury intrusion (HPMI), and scanning electron microscopy (SEM).

According to the China Oil and Gas Industry Standard SY/T 5124,[52] the vitrinite reflectance (RO) of the sample was measured by a reflective light microscope. According to the National Standard GB/T 19145,[53] total organic carbon content (TOC) was measured by a CS-230 carbon sulfur analyzer. The samples were crushed to 200 mesh, and then all powder samples were treated with hydrochloric acid at 60 °C for 24 h for decarburization and washed with deionized water to remove the residual hydrochloric acid.

The samples were analyzed by a Rigaku Ultima-IV X-ray diffractometer. According to the China Oil and Gas Standard SY/T 5163,[54] the samples were pretreated before the experiment. In order to fully disperse the minerals, the samples were crushed to less than 40 μm. The X-ray instrument scans the sample powder from 3° to 70° in 0.02° steps. The crystal structure determineed the type of minerals, and the intensity of the diffraction peak determined the level of phase content.

The LP-N2-GA experiment is based on the China Oil and Gas industry standard SY/T 6154.[55] The Beishide instrument was used in the Key Laboratory Strategic Evaluation of Shale Gas, Ministry of Land and Resources, Beijing, China. In order to complete the experiment successfully, 2 g (60–80 mesh) samples were dried in an oven at 100 °C for 24 h (lack of sample YP-7; more than 80 mesh samples were crushed in an XRD test). The dried samples were degassed in a vacuum column at 90 °C for 12 h to remove water and volatile hydrocarbons in pores. Through the first two steps, all the atmospheric moisture was discharged and then corrected by the standard sample before the experiment, and the error was less than 6%. The measurement conditions were 77.3k of liquid nitrogen, and the relative pressure (P/PO) range was 0.001–0.998. Based on the amount of nitrogen adsorption, the surface area was calculated according to the relative pressure in the range of 0.05–0.35 using the Brunauer–Emmett–Teller (BET) method.[56] Using the Barrett–Joyner–Halenda (BJH) method, the pore volume and pore size distribution parameters were obtained from the adsorption curves in the pore size range of 1.7–200 nm at the relative pressure of 0.06–0.99.[53] The specific method is described in detail in the literature.[56−59]

According to the National Standard GB/T 29171,[60] the mercury intrusion test was carried out using a Quantachrome poremaster-60 automatic high-pressure mercury porosimeter. Sample pretreatment: 4 g shale samples (1–20 mesh) were dried in a vacuum at 110 °C for 12 h and then put into the instrument for testing.

The SEM test is based on the China Oil and Gas Industry Standard SY/T 5162,[61] carried out with the FEI Quanta FEG 450 environmental scanning electron microscope. Before the experiment, the samples were polished by argon ion technology to create an artifact-free surface.

Experimental Results

Organic Geochemistry Characteristics

Maceral analysis (Table ) suggests that II2 kerogen is the main type in the sample cores, with sapropelinite content ranging from 67.3% to 69% with an average of 68% followed by inertinite accounting for 24.3%∼26.7%, with an average of 25.1%; the contents of exinite and vitrinite are less than 7%. Among the five samples tested, only YP-2 is composed of gas-prone type III kerogen with the vitrinite and inertinite in the dominant position in 100% of the maceral compositions. The change of macerals shows that the sapropelinite content of deep shale core samples accounts for the majority. At this time, the input of organic matter is mainly plankton and microorganisms. With the depth becoming shallow, terrestrial higher plants gradually input, indicating that the sedimentary water body becomes shallow gradually.

Table 1

Results of Kerogen Microscopic Analysis of the Permian Shanxi Shalesa

sample	depth	relative content of maceral groups (%)				type index	kerogen
ID	(m)	sapropelinite	exinite	vitrinite	inertinite	(TI)	type
YP-2	3556.7	0	0	40.3	59.7	–89.92	III
YP-5	3577.2	68.3	0.7	6.7	24.3	39.33	II2
YP-6	3582.3	67.3	0.3	5.3	26.7	37.17	II2
YP-8	3594.2	69	0.3	5.7	25	39.92	II2
YP-10	3600.8	68	0.7	7	24.3	38.75	II2

Note: TI = (Sapropelinite × 100 + Exinite × 50 – Vitrinite × 75 – Inertinite × 100)/100, TI > 80, 80 > TI > 40, 40 > T, I> 0, and TI < 0 indicate type I, type II1, type II2, and type III, respectively.

The maceral groups can be distinguished into four kerogen types according to the kerogen index (TI): sapropelic (I), humic-sapropelic (II1), sapropelic-humic (II2), and humic (III). The calculated results showed that the organic matter in the Permian Shanxi Shale is mainly type II2 with TI values between 40 and 0.

The organic carbon content (TOC) is between 0.37% and 7.49%, with an average value of 2.74%, indicating that the core sample is rich in organic matter and that the vitrinite reflectance (Ro) varies from 2.60% to 3.08%, with an average value of 2.9%, revealing that it is in the stage of overmature gas generation. It can be seen from Figure that the Ro value increases with the increase of depth. Due to the evolution of diagenesis and the sedimentary environment, TOC does not change significantly with depth, but the overall trend also increases. At the same time, it is found that the TOC content changes greatly in the depth of sample YP-4 (3577.2 m), indicating that the water body is relatively turbulent at this time, while the water body is relatively stable in the shallow deposition period of 3577.2 m.

Figure 2

Organic maturity and TOC contents of the Permian Shanxi Shales.

Mineralogical Composition

According to the XRD experimental data, the mineral composition of the core sample is shown in Table . The shale of Shanxi Formation is mainly composed of clay minerals with an average content of 65.2% ranging from 51.6% to 89.1%, followed by quartz with an average content of 29.8% ranging from 8.0% to 41.7%. The content of siderite and feldspar is low, with an average content of 2.8% and 1.3%. In addition, no carbonate minerals (calcite and dolomite) are detected, and pyrite only exists in YP-7 and YP-9. It can be seen that the shale clay mineral content of Shanxi Formation is higher, which is conducive to the enrichment of organic matter.

Table 2

Mineral Composition of the Permian Shanxi Shalea

	mineral composition (%)					clay minerals (%)
sample ID	clay	quartz	feldspar	siderite	pyrite	kaolinite	chlorite	illite	I/S
YP-1	56.2	39.2	1.6	3.0	0.0	22	10	51	18
YP-2	59.7	31.0	0.7	8.6	0.0	23	3	41	33
YP-3	61.7	34.4	1.5	2.4	0.0	27	6	48	19
YP-4	55.3	41.7	2.4	0.6	0.0	43	7	42	8
YP-5	64.4	32.5	2.4	0.7	0.0	63	19	15	3
YP-6	56.1	40.6	2.2	1.0	0.0	34	10	56	0
YP-7	51.6	40.4	0.9	0.0	7.1	62	5	27	6
YP-8	61.5	30.3	1.1	7.1	0	68	4	28	0
YP-9	85.7	8.0	0.0	3.1	3.2	92	3	5	0
YP-10	89.1	9.8	0.0	1.2	0.0	83	3	8	6

I/S denotes illite–smectite mixed.

In terms of clay minerals, kaolinite accounts for the largest proportion, ranging from 22% to 92%, with an average of 52%; illite takes the second place, with an average content of 32%, ranging from 5% to 56%. The average content of I/S and chlorite is 14% and 8%, respectively. From Figure , we can directly observe the change of mineral content with depth. With the increase of depth, the content of quartz gradually decreases, and the content of clay minerals increases (Figure a). No significant changes in the contents of major minerals (clay and quartz) were observed in 10 samples, indicating that the sedimentary environment did not change significantly at that time. The transformation between clay minerals changed significantly up and down with the depth of sample YP-4, and kaolinite transforms into illite (Figure b).

Figure 3

Distribution diagrams of mineral composition (a) and clay minerals (b).

Micromesoporous Parameters

Isotherm Characteristics

Macropores (>50 nm), mesopores (2–50 nm), and micropores (<2 nm) were distinguished according to the standards of The International Union of Pure and Applied Chemistry.[62] The LP-N2-GA experiment is mainly used to characterize mesopores and micropores, to reflect the gas storage capacity of rock.[63,64]

The adsorption isotherms can be classified into six types, which are shown in Figure ,[65] with each curve described in detail. It can be observed from Figure that the isotherms of the Permian Shanxi Formation shale belong to the fourth type. When P/Po is in the range of 0–0.45 MPa, monolayer adsorption occurs on the shale surface and then increases slowly, indicating that monolayer adsorption is saturated to multilayer adsorption. When P/Po is in the range of 0.45–0.9 MPa, the desorption curve is higher than the adsorption curve, which is a hysteresis loop due to the capillary condensation of mesopores. When P/Po is in the range of 0.9–1.0 MPa, the two curves rise rapidly until the pressure of water vapor is close to saturation. The phenomenon of adsorption saturation is not seen, indicating that there are some mesopores and macropores in the sample. Also the adsorption capacity of the sample is low, indicating that it contains a small amount of micropores.

Figure 4

Adsorption isotherm types. Reprinted in part with permission from ref (65). Copyright 1985 Walter de Gruyter.

Figure 5

N2 adsorption–desorption isotherms of the shale sample: adsorption (solid rhombu) and Desorption (empty rhombu).

Characteristics of Hysteresis Loops

The hysteresis loops are divided into four types corresponding to their different pore morphology (Figure ). The results show that the adsorption and desorption curves of the H1 hysteresis loop are very steep, and the relative pressure of capillary condensation is in the middle, which generally corresponds to the cylindrical hole with two ends open. The adsorption and desorption curves of the H2 hysteresis loop are quite different, similar to the “big belly” shape, which generally corresponds to the ink bottle type pores and has poor connectivity and uneven pore structure. The characteristics of the H3 hysteresis loop curve are as follows. When the relative pressure is close to the saturated vapor pressure, the adsorption curve rises suddenly, and the pores are generally wedge shaped and formed by the loose accumulation of flaky particles; the adsorption curve and desorption curve of the H4 hysteresis loop are relatively parallel and flat, which is due to the parallel pore structure in the rock, corresponding to the parallel plate pore.

Figure 6

Four types of hysteresis loops and their related pore shapes. Adapted with permission from ref (65). Copyright 1985 Walter de Gruyter.

Analysis of Figure : The hysteresis loops of Shanxi shale belong to H3 and H4 types (YP-7,8,10). Because the particle size of YP-7 sample powder is larger than other samples, the adsorption capacity is too high. It reflects that the pore types of the samples are mainly wedge shaped and parallel plate shaped. The low-pressure area (0 < P/Po < 0.45 MPa) and the adsorption and desorption curves coincide basically, which indicates that one end of the closed pore is dominant in a smaller pore size; there is a medium pressure section (0.45 < P/Po < 0.9 MPa); the adsorption curve clearly lagged behind the desorption curve; and there is a steep drop in the range of 0.48–0.52 MPa, which reflects that the sample is mainly wedge shaped and parallel plate shaped pores in larger pore size, with fracture development.

Characteristics of Pore Structure Parameters

The results of LP-N2-GA experiments for pore structure parameters are shown in Table . It can be seen that the PV of Shanxi shale is low, ranging from 0.0085 to 0.0146 mL/g, with an average of 0.0126 mL/g. The average SA is 4.282 m2/g, and the average pore size is 12.6 nm (except for YP-7).

Table 3

Pore Structure Parameters of the Permian Shanxi Shalea

sample	depth (m)	SA (m2/g)	PV (mL/g)	APS (nm)
YP-1	3551.56	5.327	0.0142	10.7
YP-2	3556.64	2.681	0.0102	15.2
YP-3	3564.11	4.533	0.0145	12.8
YP-4	3572.23	4.932	0.0154	12.5
YP-5	3577.22	5.427	0.0125	9.2
YP-6	3582.28	5.611	0.0146	10.4
YP-7	3589.14	14.790	0.0514	13.9
YP-8	3594.19	2.617	0.0085	13.0
YP-9	3599.56	5.781	0.0146	10.1
YP-10	3600.83	2.204	0.0097	17.6

APS, average pore size; SA, BET surface area; PV, BJH pore volume.

The average pore size (APS) of the sample is in the range of 9.2–17.6 nm, with an average of 12.6 nm, which is summarized from Table . The pore size distribution (PSD) is characterized by the BJH method, and its image is shown in Figure . In general, there is only one peak in the curve of shale samples, which is concentrated in the range of 2.5–4 nm, indicating that the proportion of pores in the shale sample is large. The change rate of pore volume with pore diameter increases with the increase of pore diameter. When the pore size is 60 nm, there are two changes: YP-3 and YP-4 decrease obviously and slightly, and the curve shape of other samples still rises, indicating that a certain number of macropores are developed in shale samples at the same time. It can be clearly observed that the pore diameters of YP-1, YP-3, and YP-4 samples are significantly higher than those of other samples in the range of 10–100 nm, indicating that the mesopores and macropores of these three groups of samples are developed compared with other samples.

Figure 7

Pore size distribution (BJH) of the shale sample.

Macropore Parameters

Different from the LP-N2-GA method, the HPMI method can measure the pore size distribution characteristics of shale samples with large pores,[66] which can characterize more micropore throats, thus reflecting the seepage capacity of rocks. According to the high-pressure mercury injection curve (Figure ), the mercury injection curve of shale samples can be divided into two stages: in the low-pressure stage (<10 MPa), with the increase of pressure, the mercury injection saturation increases slowly, indicating that the shale samples have macropores (>50 nm), and in the high-pressure stage (>10 MPa), the mercury injection saturation increases linearly with the increase of pressure until the maximum pressure, which reflects the high pressure. This indicates that there are a lot of mesopores (<50 nm) in shale samples. The mercury removal curves of YP-1, 2, 3, and 4 were significantly higher than those of YP-5, 6, 7, 8, 9, and 10, indicating the development of the macropore ratio of the first four samples and the last six samples, which was also consistent with the results of the nitrogen adsorption curve.

Figure 8

Capillary pressure curves of the shale samples for mercury injection (solid rhombu) and mercury ejection (empty rhombu).

The relationship between the pore throat distribution range and the pore throat distribution frequency is shown in Figure , showing the pore throat size distribution of shale samples. The pore throat diameter of shale samples develops in the range of 5–130 nm, and the peak appears in the range of 8–11 nm, indicating that there are a large number of mesopores and macropores in Shanxi shale samples. The corresponding pore characteristics were observed in the scanning electron microscope (Figure ). The results show the diagenetic contraction joint (a); the structural stress joint (b); the organic matter shrinkage joint (c); and the pyrite inner pore (d).

Figure 9

Distribution range of pore throat versus pore throat distribution frequency.

Figure 10

SEM images of Shanxi Formation shale: diagenetic shrinkage fracture (a); tectonic stress fracture (b); organic matter shrinkage fracture (c); and intergranular pore of pyrite (d).

FHH Fractal Dimension

The irregularity of the pore surface and the complexity of the pore structure play an important role in gas adsorption and desorption. The Frenkel–Halsey–Hill (FHH) model can quantitatively characterize the complexity of shale pores, which has been used by many scholars.[67−69] The expression is as followswhere V is the volume of adsorbed gas at different relative pressures (P/Po); Po is the saturated vapor pressure of the gas; a is the constant; b is the slope of the straight line; and D is the constant.

It can be seen from Figure that the magnetic hysteresis loop appears in the isotherm near the relative pressure of 0.45, which reflects different adsorption mechanisms and can be used to divide the fractal range. In particular, samples 2, 8, and 10 have no nitrogen adsorption capacity when the relative pressure is less than 0.035, and the fractal dimension of sample YP-7 in this range is 1.1688 < 2, which indicates that the pore diameter in this region is less than 2 nm and that it is not classified under relative pressure, and it is observed that the contribution of nitrogen adsorption capacity to the total adsorption capacity is very small under relative pressure, which is not further discussed. Therefore, the two intervals are Section A (0.035 < P/Po < 0.45), whose fractal dimension D1 is used to characterize the regularity of the pore surface, and Section B (P/Po > 0.45), whose fractal dimension D2 can characterize the complexity of the pore structure.[67]

Figure and Table show the fractal curves and dimension of ten shale samples from the Shanxi Formation, respectively. The correlation coefficient is more than 0.99, indicating that the equation has a good fitting relationship. In the pore fractal, the fractal dimension is between 2 and 3. The larger the fractal dimension is, the more complex the pore structure is and the stronger the heterogeneity is. D1 varies between 2.4342 and 2.6553 with an average of 2.5582, and D2 changes from 2.5553 to 2.7102 with an average of 2.6255. In general, the pore structure of the sample is complex, heterogeneous, and strong. D2 is larger than D1, which indicates that the complexity of the pore internal structure is greater than that of the pore surface structure.

Figure 11

Fractal dimensions of the Shanxi shale sample.

Table 4

Fractal Dimensions D1 and D2 and Correlation Coefficient R2

	section A			section B
sample ID	equation	D1	R2	equation	D2	R2
YP-1	y = −0.3845x + 0.6247	2.6155	0.9981	y = −0.4055x + 0.6048	2.5945	0.9991
YP-2	y = −0.4579x – 0.0604	2.5421	0.9944	y = −0.4206x – 0.0231	2.5794	0.9981
YP-3	y = −0.4204x + 0.4667	2.5796	0.9975	y = −0.3592x + 0.5356	2.6408	0.9848
YP-4	y = −0.4235x + 0.5478	2.5765	0.9986	y = −0.3978x + 0.5602	2.6022	0.9988
YP-5	y = −0.3971x + 0.6347	2.6029	0.9999	y = −0.2898x + 0.7046	2.7102	0.9923
YP-6	y = −0.4003x + 0.6713	2.5997	0.9994	y = −0.3305x + 0.7148	2.6695	0.9985
YP-7	y = −0.3447x + 1.6429	2.6553	0.9999	y = −0.4072x + 1.5389	2.5928	0.9914
YP-8	y = −0.5629x – 0.1198	2.4371	0.9953	y = −0.3861x – 0.0327	2.6139	0.993
YP-9	y = −0.4345x + 0.7053	2.5655	0.9993	y = −0.3185x + 0.7764	2.6815	0.9985
YP-10	y = −0.5658x – 0.2796	2.4342	0.9949	y = −0.4447x – 0.2171	2.5553	0.9984

Discussions

Relationships between TOC Clay and Quartz Contents

The changes of mineral composition can reflect different sedimentary environments and diagenetic evolution. As shown in Figure , the total TOC of deep shale in the Shanxi Formation has no obvious correlation with clay minerals and quartz content. However, comparative marine facies of the Dalong Formation, Longmaxi Formation, and Niutitang Formation and the TOC and clay minerals are significantly negatively correlated and positively correlated with quartz.[69,−73] Compared with the transitional Shanxi Formation, the TOC is positively correlated with clay minerals and negatively correlated with quartz.[74,75]

Figure 12

Relationships between TOC and clay (a) and quartz (b).

Different sedimentary environments lead to the different relationship between TOC content and mineral composition. Organic matter in marine shale mainly comes from planktonic algae far away from land, which is not conducive to the input and enrichment of land clay minerals.[76−78] Therefore, TOC is negatively correlated with clay mineral content. In addition, because the lower part of the water body is in a strong reducing environment, the quartz content in marine shale is high, and most of them are biogenic, which is positively correlated with TOC.[79−82] Shallow transitional shale is due to the input of terrigenous clay minerals and the lack of biological quartz, and the relationship between TOC, clay minerals, and quartz is opposite to that of marine shale.

The shale samples were deposited in the transitional facies. There is no obvious correlation between TOC, clay minerals, and quartz, which indicates that the Shanxi Formation is unstable in the sedimentary period. The input of land-based clay minerals is not strong, and there is also some quartz in shale. Therefore, the presence of quartz and detrital quartz in Shanxi shale leads to no obvious correlation between TOC and quartz.

Relationships between Pore Structure Parameters

Figure shows the relationship between different pore structure parameters of Shanxi shale. APS was negatively correlated with SA (Figure a) and PV (Figure b), with correlation coefficients of 0.801 and 0.3824, respectively. It is well known that mesopores play an important role in SA and PV.[83,84] The negative correlation between APS and SA and PV indicates that the shale with smaller APS has more mesopores, which is consistent with the previous research results of highly mature shale.[85−87] There is a significant positive correlation between PV and SA (R2 = 0.7824) (Figure c), which is consistent with the correlation between marine continental transitional facies and marine shale.[88−90]

Figure 13

Relationships between surface area and average pore size (a), pore volume and surface area (b), and average pore size (c).

Relationships between TOC, Minerals, and Pore Structure

The relationship between TOC, mineral composition, and SA and PV of shale is shown in Figure . There is a significant negative correlation between pore structure parameters and TOC (Figure a), and there is no correlation with quartz, kaolinite, and illite (Figure b, c, d).

Figure 14

Relationships between surface area, pore volume, and TOC (a), quartz (b), kaolinite (c), and illite (d).

There was a significant negative correlation between TOC and SA and PV, and the correlation coefficients were 0.5081 and 0.7099, respectively. This not only indicates that the contribution of organic matter to SA and PV is very small and may even block the pore space but also indicates that the organic pores of Shanxi shale dominated by type II2 kerogen are not developed. Previous studies have also confirmed that the organic pores in shale dominated by type III and type II kerogen are poorly developed. Only a few organic pores can be seen under the scanning electron microscope (Figure a,b). Due to the large burial depth, the organic pores are easily damaged by compaction. The organic matter fibrosis leads to the destruction of organic pores, and the number of organic pores is significantly reduced. The matrix asphalt in shale can also fill the organic pores, resulting in the blockage of organic pores, reducing its contribution to SA and PV;[41] the fractures at the edge of organic matter are obviously developed (Figure c) because hydrocarbon generates from the organic matter which can result in the shrinking of OM. The organic matter shrinks to form microstructures. This is contrary to the correlation of marine Longmaxi Formation and Niutitang Formation shale.[74]

Figure 15

SEM images of organic pores in Shanxi shale. (a, b) A small amount of organic pores; (c) no organic pores, with organic matter edge fractures developed; and (d) organic matter and clay minerals are mixed and filled between particles, without organic pores.

The relationship between shale mineral composition and pore structure parameters can reflect the degree of pore development related to mineral composition to a certain extent. Quartz, kaolinite, and illite have no obvious correlation with SA and PV, indicating that the contribution of shale minerals in deep Shanxi Formation to SA and PV is very small. The kaolinite of shallow buried transitional Shanxi Formation is negatively correlated with SA and PV, and I/S is positively correlated with SA and PV. The illite of marine Niutitang Formation and Longmaxi Formation is positively correlated with SA and PV, and quartz is not significantly correlated with SA and PV.[41,91] The correlation between D1 and TOC and mineral components is worse than D2, indicating that D2 is more closely related to pore structure characteristics.

Previous studies have shown that clay minerals in shale usually contain nanopores, which can provide a certain adsorption site and storage space for shale gas.[74,92] Illite has a certain contribution to SA and PV in marine and transitional facies shale, but there is no obvious correlation between illite and SA and PV in deep Shanxi Formation shale. Although illite easily produces wedge-shaped pores, it also easily fills organic pores, which is not conducive to pore development.[93,94] Shale samples are deeply buried, and clay minerals and organic matter will fill and plug the primary pores; therefore, the mineral content of quartz, kaolinite, and illite has little contribution to SA and PV. Kaolinite and illite have a weak negative correlation with SA and PV, indicating that illite has more micropores than kaolinite, which is consistent with previous research results.[95]

Relationships between Fractal Dimensions and Pore Structure

The fractal dimensions D1 and D2 reflect the surface roughness and the complexity of the internal structure of pores, respectively. The larger the proportion of micropores, the larger the SA, the more complex the pore structure, and the larger the corresponding fractal dimension.[96]Figure shows the relationship between pore structure parameters and the fractal dimension of shale. The results show that the fractal dimension is positively correlated with SA and PV (Figure a,b) and negatively correlated with APS (Figure c). Previous studies have also confirmed this relationship.[97,98]

Figure 16

Relationships between fractal dimension and surface area (a), pore volume (b), and average pore size (c).

The pore structure parameters measured by nitrogen adsorption–desorption experiments can basically represent the pore structure of the total pores in these shale samples. The correlation coefficients of D1 with specific surface area and pore volume were 0.7534 and 0.6971, respectively, which were larger than D2 (0.5384 and 0.1941), indicating that the larger D1 was, the larger the specific surface area and pore volume were. Therefore, D1 can more effectively reflect the development degree of micropores in shale samples.

Relationships between TOC, Minerals, and Fractal Dimensions

Discussion on the influencing factors of fractal dimension is helpful to further understand the formation mechanism and influencing factors of shale heterogeneity.[99] Different mineral contents have different effects on pore heterogeneity. As can be seen from Figure , the fractal dimension D1 is negatively correlated with TOC and positively correlated with quartz, with correlation coefficients of 0.3294 and 0.3386, respectively; D2 has no obvious correlation with TOC and quartz (Figure a,b); the fractal dimension has no obvious correlation with kaolinite and illite (Figure c,d). This is inconsistent with the previous research results on the correlation between mineral composition and fractal dimension of marine and transitional shale.[69,71,99]

Figure 17

Relationships between fractal dimensions and TOC (a), quartz (b), kaolinite (c), and illite (d).

The fractal dimension of shale samples is negatively correlated with TOC content. Generally speaking, the more developed the organic pores are, the larger the specific surface area is, and the more complex the pore structure is. The maturity of organic matter in Shanxi Formation shale reached the overmature stage, and the degree of organic matter carbonization was high, resulting in the poor development of organic pores. Moreover, the buried depth of the shale sample is large, and the original organic pores collapse due to strong compaction. Therefore, TOC is negatively correlated with fractal dimension, which further confirms the conclusion of Section . The fractal dimension of the shale of Dalong Formation is positively correlated with TOC content, which may be related to the maturity and burial depth of organic matter.[41] The rougher pore surface has no obvious correlation with D2. The fractal dimension of the shale of Dalong Formation, a shallow transitional facies, is negatively correlated with quartz content.[100] Quartz is affected by brittleness, dissolution, and secondary expansion and has poor correlation with fractal dimension. The fractal dimensions D1 and D2 have no obvious correlation with kaolinite and illite, indicating that the pores related to clay minerals are not dominant in Shanxi shale, and clay minerals have strong plasticity. Under compaction, the pores are filled or collapsed by organic matter.

In general, the correlation between D1 and TOC and mineral composition is better than D2, indicating that the correlation between fractal dimension D1 and pore structure characteristics is closer in deep shale.

Comparison of Pore Structure between Deep and Shallow Shale

The organic matter type of shallow shale of Shanxi Formation in Ordos Basin is mainly III. The TOC is between 2.11–2.53; the Ro average range is 1.25–2.58%; clay mineral content is 51.7–64.71%; quartz content is 32.7–43.7%; pore volume is mainly mesoporous and macroporous; micropores are few; and there is a parallel plate shape.[101−105] Compared with the deep shale in this paper, TOC and Ro are lower, but the content of mineral composition is a little different.

The study of deep shale pores is mainly carried out around the Sichuan Basin. With the increase of depth, organic pores, inorganic pores, and microfractures are increasing, and an effective pore network is formed between them,[106] which is conducive to the migration of shale gas. The high overlying formation pressure has a limited effect on the pores of deep shale, and some deep shale still retains large pore size and regular pore morphology, which is conducive to the preservation of micropores. The developed natural fractures are conducive to the enrichment of shale gas.[107,108] The fractal dimension is 2.72–2.92, in which D2 is greater than D1, and the complexity of the pore structure is greater than that of the pore surface,[109,110] which is similar to the research in this paper. The Shanxi Formation in Ordos Basin is widely distributed, is deeply buried in the study area, and has good pore structure characteristics, and microfractures are developed, which is of positive significance for the preservation and enrichment of shale gas.

Conclusions

Through geochemical analysis, low-pressure nitrogen adsorption–desorption, high-pressure mercury injection, and scanning electron microscope experiments and FHH theory, the pore structure and classification characteristics of deep shale in Shanxi Formation were studied. The following conclusions are reached:

Deep shale in Shanxi Formation is deeply buried with high TOC content. It is mainly composed of clay minerals, and the organic pores in the type II2 kerogen in transitional Shanxi shale are not developed, which is opposite to the marine shale dominated by quartz and abundant organic pores and also distinguishes it from the pore development of clay minerals in shallow and medium transitional shale.

The main pore types of Shanxi deep shale are intergranular pore, microfracture, and organic matter shrinkage fracture. The pore size is mainly in the range of 2.5–4 nm and 8–11 nm, and the main pore shape is wedge and parallel plate. The original pores of deep transitional shale were compacted under the action of overlying formation pressure, and the role of minerals in pore structure is not obvious. Although organic pores are immature, organic matter carbonizes in the overmature stage, and a large number of organic marginal fractures are produced in the hydrocarbon generation stage, which plays a dominant role in the pore system of shale samples. This is also confirmed by SEM experiments.

According to the fractal FHH model, the fractal dimension of Shanxi shale is relatively large. The average values of fractal dimension D1 of the pore surface and fractal dimension D2 of the pore structure are 2.5582 and 2.6255, respectively. D1 and D2 are negatively correlated with TOC, and the correlation between mineral components is weak, which verifies that organic pores are not developed and that minerals have little contribution to SA and PV. D1 > D2, which shows that the complexity of the pore structure is greater than that of the pore surface.

The reservoir space of deep Shanxi Formation shale is mainly affected by the TOC content, burial depth, and hydrocarbon generation. Compared with previous studies, the pore characteristics of deep shale are a large proportion of mesopores, small specific surface area, large average pore size, and more developed microfractures. A large number of microfractures were observed by a scanning electron microscope. Compared with marine and midshallow transitional shale, mineral pores had little contribution to SA and PV. Therefore, more attention should be paid to the study of microfractures in the later exploration and development process.