J. Cent. South Univ. (2019) 26: 925-937

DOI: https://doi.org/10.1007/s11771-019-4061-3

Effects of caprock sealing capacities on coalbed methane preservation: Experimental investigation and case study

TIAN Fu-chao(田富超)^{1, 2}, LIANG Yun-tao(梁运涛)^{1, 2}, WANG De-ming(王德明)², JIN Kan(金侃)¹

1. School of Safety Engineering, China University of Mining and Technology, Xuzhou 221116, China;

2. State Key Laboratory of Coal Mine Safety Technology, China Coal Technology & Engineering Group Shenyang Research Institute, Fushun 113122, China

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2019

Abstract: Caprocks play an important role in the trapping of coalbed methane (CBM) reservoirs. To study the sealing capacities of caprocks, five samples with different lithologies of Neogene clayrock, Paleogene redbeds, Permian sandstone, Permian mudstone and Permian siltstone were collected and tested using experimental methods of microstructure observation, pore structure measurement and diffusion properties determination. Results indicate that with denser structures, lower porosities, much more developed micropores/transition pores and higher pore/throat ratios, mudstone and siltstone have the more ideal sealing capacities for CBM preservation when comparing to other kinds of caprocks; the methane diffusion coefficients of mudstone/siltstone are about 6 times higher than sandstone and almost 90 times higher than clayrock/redbeds. To further estimate the CBM escape through caprocks, a one-dimensional CBM diffusion model is derived. Modeling calculation result demonstrates that under the same thickness, the CBM sealing abilities of mudstone/siltstone are almost 100 times higher than those of clayrock/redbeds, and nearly 17 times higher than sandstone, which indicates that the coal seam below caprocks like clayrock, redbeds or sandstone may suffer stronger CBM diffusion effect than that below mudstone or siltstone. Such conclusion is verified by the case study from III3 District, Xutuan Colliery, where the coal seam capped by Paleogene redbeds has a much lower CBM content than that capped by the Permian strata like mudstone, siltstone and sandstone.

Key words: caprock; sealing capacity; coalbed methane; preservation

Cite this article as: TIAN Fu-chao, LIANG Yun-tao, WANG De-ming, JIN Kan. Effects of caprock sealing capacities on coalbed methane preservation: Experimental investigation and case study [J]. Journal of Central South University, 2019, 26(4): 925–937. DOI: https://doi.org/10.1007/s11771-019-4061-3.

1 Introduction

Coalbed methane (CBM) is a self-generated and self-reserved unconventional natural gas resource that is generated in the coalification process and accumulated in coal reservoirs [1], which has always been treated as outburst and explosive hazards in mining [2, 3]. The accumulation of CBM is actually the process of CBM preservation, and the amount of CBM preserved in reservoirs depends on several factors, such as coal rank, burial depth, lithologies of the surrounding rocks, local geologic anomalies [4, 5]. Generally accepted theory considers that CBM pools are mostly accumulated by adsorption mechanism, thus a conventional trapping mechanism is not very much required (namely the requirements of caprocks for CBM reservoirs are weaker than conventional gas reservoirs) [6]. However, as research continues, the importance of the caprocks has been more and more revealed [7].

Recently, models established for CBM potential evaluations or carbon dioxide geological sequestrations gradually began to regard the caprock properties as key parameters [8].

From a viewpoint of geology, the coal measure strata underwent several tectonic movements after the formation of coal reservoirs, resulting in several subsidences and uplifts and eventually reached nowadays positions [4]. During the uplift stages, along with the upthrow, the coal measure strata always suffered from a long period of denudation and sometimes might even be exposed to the atmosphere, which led to a large quantity emission of the thermogenic methane that was generated through plutonic metamorphism [9, 10]. Meanwhile, when the coal seams were lifted to a near surface elevation, the methanogens introduced by meteoric water can decompose the organic matter in coal and generate secondary biogenic gas [11], which could significantly re-charge the coal seam. Studies by QIN et al [12] demonstrated that the origins of CBM in China consist of both biogenic and thermogenic gasses. Therefore, the role that caprocks play in CBM preservation became much more important because if the sealing capacities of the overlying caprocks were poor, the late generated secondary biogenic gas in coal reservoirs would easily escape. It has been reported that, in the coal basins of Silesian (Poland) [13], Sydney (Australia) [14], Qinshui (China) [15] and Ordos (China) [16], higher gas contents seem to be concentrated in the areas where coal reservoirs are overlain by impermeable overburden, indicating that the CBM preservation conditions are strongly inhomogeneous due to the features of caprocks.

Generally, the effects of caprocks’ sealing capacities on CBM preservation are closely related to their formation (sedimentary) environments [17–19] and can be primarily reflected by the lithologies and physical properties of the caprock strata [16]. Poorer sealing capacities of the caprocks can lead to an easier loss of CBM, and thus the gas contents of coal reservoirs will be lower [20]. However, even though the sealing capacities of caprocks have significant impacts on collieries’ CBM occurrences and gas disasters, corresponding studies of the effects of caprock sealing properties on CBM preservation are still seriously lacking. SAGHAFI et al [21] only measured the differences of permeabilities and diffusivities between claystone and sandstone. ZHANG et al [22] compared the difference in sealing properties among mudstone, siltstone, sandstone and limestone but did not provide any data. JIN et al [7] experimentally investigated the CBM sealing capabilities of different caprocks but didn’t provide any further discussion.

The Xutuan Colliery is located in the southern region of the Huaibei Coal Basin, northern Anhui Province of China, covering an area of 52.59 km². The main commercial coal seams of the colliery are the #3₂, #7₂ and #8₂ coal seams. Due to the unique tectonic movements near the coalfield, there are different caprock groups in the southern/northern flanks of the III3 District, Xutuan Colliery respectively, where the caprocks in the southern flank are mainly composed by Quaternary unconsolidated formation, Neogene clayrock and Paleogene redbeds while in the northern flank are composed by Quaternary unconsolidated formation, Neogene clayrock and Permian strata (mudstone, siltstone and sandstone). Additionally, due to the fact that the CBM origin in Huaibei Coal Basin has been proven as a mixture of thermogenic and biogenic gasses (47.80%–47.98% of the CBM is biogenically generated) [9, 23]. Therefore, the III3 District, Xutuan Colliery can provide an excellent case for the comparative study of the effects of caprocks’ lithologies on CBM preservation.

This article takes the CBM preservation in III3 District, Xutuan Colliery, Huaibei Coal Basin as a typical case. Through a combining use of experimental investigations (microstructure observation, pore structure measurement, diffusion properties determination), modeling calculations and case study, the CBM sealing capacities of different kinds of caprocks are comprehensively evaluated. Results of this study can somehow reveal the CBM preservation mechanism and can be used to guide the CBM potential estimates.

2 Sampling and methods

2.1 Sample collection

To study the effect of caprocks’ sealing capacities on CBM preservation, five kinds of caprocks were sampled from the overlying strata of the #3₂ coal seam, Xutuan Colliery by surface drilling, including Neogene clayrock, Paleogene redbeds, Permian sandstone, Permian mudstone and Permian siltstone. The Quaternary sample was failed to collect because of its highly unconsolidation. Once a sample was obtained, the sample was sealed and immediately transported to the laboratory for the following sample preparations and tests.

Some intact cores of caprock samples were selected to make φ25 mm×50 mm standard samples for methane diffusion experiments, while other fragments were further broken and sieved to the particle size of 10–12 mm and 3–6 mm for the microstructure observation and pore structure analyses, respectively.

2.2 Experimental methods

Researches revealed that the dissipation of CBM has three paths [24]: 1) free gas dissipates by overcoming capillary pressure of sealing rocks;2) dissolved gas diffuses by concentration difference; 3) hydrodynamic loss is caused by water flushing. The former two paths are both significantly influenced by the properties of caprocks. Thus, the gas preservation mechanism of caprocks that affects the sorption-type CBM reservoirs is that the caprocks with favorable sealing capabilities can effectively prevent the gas dissipation through the overburden of the coal seam [16].

To comprehensively evaluate the sealing capacities among different caprock samples, scanning electron microscope (SEM, Quanta 250, FEI, USA) was introduced to observe the microstructures of the caprock samples. Then, pore structures and gas diffusion properties of the samples were analyzed to quantitatively characterize the sealing capacities of different caprocks. The pore structures were tested by the mercury intrusion porosimetry (MIP) using a PoreMaster 33 automated mercury intrusion porosimeter (Quantachrome Instruments, USA), and the data were modeled to the Washburn Equation:

                           (1)

where p_c is the capillary pressure (MPa); σ is the surface tension of Hg (dyn/cm²); θ is the contact angle between Hg and coal surface (°); and r is the pore/throat radius (nm).

The gas sealing capabilities of the caprocks were evaluated by the methane diffusion coefficients of the samples, with a KDKX-II block coal diffusion coefficient tester (Kedi Petroleum Instrument, China) using a Fick’s first law based free gas diffusion method [25]. According to Fick’s law and the law of conservation, the diffusion coefficients can be derived as follows:

                         (2)

whereD is the diffusion coefficient (m²/s); ΔC_f0 is the concentration difference of the hydrocarbon gas in two gas chambers at the initial moment (%); ΔC_fi is the concentration difference of the hydrocarbon gas in two gas chambers at i moment (%); A is the sectional area of the sample (cm²); V₁, V₂ are the volumes of two gas chambers respectively; l is the length of the sample (cm); t₀ is the initial moment of diffusion (s); t_i is the i moment of diffusion (s).

3 Results and discussion

3.1 SEM observation

The microstructures of caprock samples were analyzed by observing the fresh surfaces of the rock sections using the SEM technology. The SEM images of different caprock samples are shown in Figure 1. After a magnification of 2000 times, the microstructural images of Neogene clayrock, Paleogene redbeds and Permian sandstone all show a large number of micron-size pores, resulting in high connectivities and developed pore structures. However, the pore development degree of sandstone is less remarkable than the other two samples, but the fractures developed in sandstone also can enhance the permeability of this rock. Thus, the CBM sealing capabilities of these three rocks are weak, especially for clayrock and redbeds.

Conversely, the Permian mudstone and Permian siltstone whose matrixes are mainly composed of flaggy and schistose grains are much denser. Even under a magnification of 10000 times, no obvious pores or fractures can be found. Therefore, with less developed pore structures, these two types of caprocks favor the sealing of CBM.

The formation of caprocks’ pore structures is closely related to the diagenesis effects of the rocks. For clayrock and redbeds, both rocks are young strata whose burial depth is never deeper than 1000 m in the study area. Therefore, the compaction and cementation of the rocks are slight, leading to looser structure with more pores remained. Moreover, eluviation effects would further lead to the development of induced porosity, and the high content of fragile minerals (quartz, muscovite and calcite) in redbeds would also benefit the conservation of pore space.

Figure 1 SEM images of different caprock samples:

For sandstone, mudstone and siltstone, the maximum burial depths after coal seam deposition are nearly 3000 m, which result in high degree of compaction. However, the tectonic movements induced fractures and recrystallization of quartz may somehow contribute to secondary pore volume in fragile sandstone. But for mudstone and siltstone, under a high degree of compaction, their fine plasticity clastic grains will result in much denser microstructures. On the whole, the sealing capacities of old strata (Permian mudstone, siltstone and sandstone) which underwent higher degree of compaction are better than the newly deposited strata (Neogene clayrock and Paleogene redbeds).

3.2 MIP pore structure

As one of the most popular methods in pore structure characterization, the MIP method provides lots information concerning the pore structures in samples, including pore volumes, porosities, specific surface area, pore size distributions (PSDs), etc. To better analyze the pore structures quantitatively, the Hodot pore size classification method for coal/rock [7] was introduced and applied as: micropores (<10 nm), transition pores (10–10² nm), mesopores (10²–10³ nm) and macropores (>10³ nm).

3.2.1 Mercury intrusion/extrusion curves

The mercury intrusion/extrusion curves of different caprock samples are illustrated as Figure 2. From the morphologies of the MIP curves (widths of hysteresis loops, shapes of injection and ejection curves), some qualitative informations of the samples’ pore structures can be acquired. For Neogene clayrock and Permian redbeds samples, their MIP curves both exhibit broad hysteresis loops and large volume differences between their intrusion and extrusion curves, which indicate that transition pores (pore size: 10–10² nm) constitute a large ratio of the total pore volume and the pores are well connected. The shape of mercury extrusion curves shows slightly horizontal/convex shape at first and then concave, demonstrating the existence of open pores and semi-closed pores [7]. The morphology analyses of the MIP curves of clayrock and redbeds samples are quite consistent with those from the SEM observations, indicating that the structures of Neogene clayrock and Permian redbeds are looser, and their CBM sealing capacities are weak as well.

Figure 2 Mercury intrusion/extrusion curves of caprock samples

For the MIP curves of sandstone sample, at low intrusion pressure, a large intrusion volume of mercury can be easily found, which demonstrates the existence of abundant macropores. The mercury extrusion curve of sandstone is slightly convex, which is similar to clayrock and redbeds samples, indicating the existence of open and semi-closed pores as well [7]. Besides, although the total pore volume of sandstone is smaller than clayrock and redbeds, the width of hysteresis loop and the volume difference between intrusion and extrusion curves are still great when comparing to mudstone/siltstone, which means that transition pores and mesopores constitute a large ratio of the total pore volume and the pores are well connected. With this type of pore structure, sandstone do have some certain ability in sealing the CBM; however, the sealing capacity of this rock is not good enough and can not be treated as one of the ideal kinds of caprocks.

For mudstone and siltstone, due to the extremely small intrusion volume that almost is beyond the testing capacity of the instrument, the mercury curves of these two samples are rough. And at low intrusion pressure, along with the increasing of pressure, the rise of intrusion curve is slow, which indicate the absence of macropore and mesopore. The extrusion curves of mudstone and siltstone descended sharply in the high-pressure stage; the widths of hysteresis loops and the volume differences between samples’ intrusion/extrusion curves are low; all of these phenomena indicate that open pores are rare in the sample and these pores are bad connected. This type of pore structure is not conducive to gas diffusion, but favors the preservation of CBM.

3.2.2 Pore size distribution

The pore size distribution of the caprock samples is shown in Figure 3 and the data calculated from MIP experiments are listed in Table 1.

A comparison among different caprocks indicates that although Paleogene redbeds sample is not the shallowest collected sample, the porosity of redbeds (17.8660%) is the highest in all these five samples, which may be attributed to the pores induced by the eluviation effect and the rock skeleton made of fragile minerals. The porosity of Neogene clayrock (15.1245%) is just slightly lower than that of redbeds, thus both these two caprocks have loose structures. Comparing to clayrock and redbeds, the porosity of sandstone sample (6.1258%) is much smaller, which is just about 40% of that in clayrock or redbeds. However, the porosities of mudstone (1.5231%) and siltstone (1.4451%) are even smaller, which only count for about 25% of the sandstone sample. Porosities of the caprocks represent the pore volumes in the rocks, which provides the passageway for CBM migration. With higher porosity, the CBM sealing capacity of the caprock will be weakened.

Figure 3 Pore size distribution of caprock samples (MIP based)

Furthermore, the PSDs of clayrock and redbeds are mainly concentrated in transition pores (36.86% and 65.41%, respectively) and mesopores (51.31% and 25.22%, respectively). Whilst the PSD of sandstone is dominated by macropores (66.18%) and mesopores (14.71%). In contrast, the PSDs of mudstone and siltstone are concentrated in micropores (31.25% and 50%, respectively) and transition pores (43.75% and 28.57%, respectively). Generally, the transition pores in coals and rocks mainly constitute the space for capillary condensation and gas diffusion, while the mesopores and macropores constitute the slow/fast penetration intervals. The different PSDs of different samples indicate that caprocks, clayrock, redbeds and sandstone are favorable for gas diffusion and penetration, while mudstone and siltstone favor the sealing of CBM.

3.2.3 Pore/throat ratios

To better describe samples’ porous system, network models are often applied to charactering the porous solids, treating porous solids as a network of empty chambers (so called pores) interconnected by a network of smaller channels (so-called throats). Based on the network models and geometrical considerations, the intrusion curve of MIP result could be related to the distribution of pore constrictions (throats), whereas the extrusion curves could be related to the shapes of cavities beyond the constrictions (pores). The relationship between MIP extrusion and intrusion curves, which is also known as pore/throat ratios, has been widely used as a key parameter in characterizing the pore textural properties of the samples. Recent studies also indicated that the penetration ability of the pore system not only relates to the pore volume of the sample, but also relates to the configuration features of the pore and throat (pore/throat ratios); higher pore/throat ratios mean the lower penetration abilities of the caprocks while lower pore/throat ratios (closer to 1) indicate higher penetration abilities. The pore/throat ratios of different caprocks are illustrated in Figure 4.

From Figure 4, significant distinctions between mudstone, siltstone and other caprocks can be obviously noticed. At low degree of mercury saturation (S_Hg), the values of pore/throat ratios of mudstone and siltstone are much larger than those of clayrock, redbeds and sandstone, demonstrating that mudstone and siltstone have lower penetration abilities. The pore/throat ratios curves of sandstone exhibit the tendency of first increasing and then decreasing, which is mainly due to the high pore volumes of macropores. In a word, the pore/throat ratios of the caprock samples also confirm that clayrock, redbeds and sandstone are favorable for gas diffusion in caprocks, whilst mudstone and siltstone favor the sealing of CBM.

Table 1 Porosities and pore volume distributions of caprock samples

Figure 4 Pore/throat ratios of caprock samples

3.3 Methane diffusion coefficient

Gas diffusion, which is controlled by the diffusion coefficient of the porous medium and driven by the difference in hydrocarbon concentration between reservoir and caprock, has been considered the main mechanism of gas migration in the geological system. SAGHAFI [26] indicates that the differential diffusive flows in caprocks can create significant differences in the CBM contents of coal reservoirs. Thus, the CBM sealing capabilities of caprocks can be directly evaluated by the measurement of samples’ diffusion coefficients.

The measurements of different samples’ diffusion coefficients were conducted at 30 °C under an equilibrium gas pressure of 0.5 MPa, and the testing medium was CH₄. The experimental results are listed in Table 2.

The clayrock and redbeds have similar porosities, and their pore volumes are both mainly constituted by transition pores and mesopores which are favorable for gas diffusion. Experiment results indicate that the methane diffusion coefficients of clayrock and redbeds are similar to each other, which is about 6 times higher than that of sandstone, and about 90 times higher than that of mudstone or siltstone. Therefore, as caprocks, mudstone and siltstone can seal the CBM best; while the sandstone is a little weaker; clayrock and redbeds have the weakest sealing capabilities of CBM. CBM diffuses or migrates into the clayrock or redbeds and will quickly escape due to their weak sealing properties.

Table 2 Methane diffusion coefficients of different caprocks

4 Evaluation of caprocks’ CBM sealing capabilities

4.1 Basic theory

Generally accepted theory suggests that the main mechanism of gas migration in a geological system is dominated by diffusion [27]. According to the Fick’s first law, the diffusion mass flux is roughly proportional to the concentration gradient, pointing out that under the conditions of a given temperature and pressure, the molecular diffusive flux of an arbitrary component in a binary diffusion system is in right proportion to the concentration gradient of the component, namely:

                              (3)

where J is the diffusive flux of a certain component in the direction of coordinate x (mol/(m²·s)); D is the diffusion coefficient (m²/s); C is the concentration of diffusion component (mol/m³).

4.2 CBM escape through caprocks

Referring to different geological conditions, HUANG et al [28] summed up the patterns of gas diffusion through caprocks into four prototypes. As a typical self-generated and self-reserved unconventional natural gas pool [1], the CBM diffusion pattern through caprocks can be simplified as Figure 5. In this prototype, caprock is a non- hydrocarbon source rock of the gas pool and at the initial moment we presume that the methane concentrate in caprock is zero.

Figure 5 Prototype and physical model of CBM escape process

Based on the physical model in Figure 5 and from the phenomenological flux equation (Eq. (3)), the diffusion equation can be derived. For a constant diffusivity, the diffusion equation for one- dimensional diffusion takes the following form:

                             (4)

where C is the concentrations of hydrocarbon (mol/m³); t is the diffusion time (s); z is the diffusion distance (m). Under the conditions in this case, the initial and boundary conditions of Eq. (4) can be described as follows:

Initial condition:

                       (5)

Boundary condition:

                              (6)

Solving Eq. (4), the concentration variation as a function of x and t can be solved from the diffusion equation, and we obtain the diffusion effect induced methane concentrate distribution in caprock as:

     (7)

And the amount of gas escape through the caprock in time period t is:

 (8)

where Q is the diffusion amount through per unit caprock area (m³/m²); L is the thickness of the caprock (m); C₀ is the methane concentrate in coal seam (mol/m³); C₁ is the methane concentrate in permeable bed (mol/m³).

For most CBM preservation situations, the permeable bed in Figure 5 actually represents the Quaternary unconsolidated formation, within which the gas permeability is very high and the ground water in which will also take away the CBM diffused into it immediately. Therefore, to further simplify the physical model, we can assume the methane concentrate in the permeable bed to be zero, namely C₁=0, then Eq. (8) can be re-written as:

         (9)

Assume that the methane concentrate in coal seam and the diffusion coefficients of the caprocks both keep constant during the geological time, then the relationship between CBM escape amount and caprocks’ sealing capacities can be discussed. The basic parameters of the study coal seam (#3₂ coal seam) are list in Table 3.

According to the porosity of the #3₂ coal seam (6.42%) and assigning the gas pressure of the coal seam as 0.74 MPa, the methane concentrates in coal seam (C₀) can be calculated as 24.075 mol/m³. Substituting the corresponding parameters into Eq. (9), the correlations among CBM escape amount, caprock thickness and diffusion time can be obtained as Figure 6. Overall, under the same thickness, the higher diffusion coefficient of the caprock will lead to the larger escape amount of CBM. Thus, as caprock, the redbeds have the weakest sealing capacity among all these five kinds of caprocks while the mudstone and siltstone favor the sealing of CBM best. In addition, with the increase of the caprock thickness, the gas escape amount decreases in a power function tendency.

Table 3 Proximate analyses and adsorption constants of the #3₂ coal seam, Xutuan Colliery

Figure 6 Correlation among CBM escape amount, caprock thickness and diffusion time:

With the same thickness of caprock (assume a thickness of 20 m), the correlations between gas escape amount and escape time are illustrated in Figure 7. The CBM sealing abilities of mudstone and siltstone are almost 100 times higher than that of clayrock or redbeds, and nearly 17 times higher than that of sandstone. This result is consistent with that of methane diffusion coefficient measurement, which indicates that the methane diffusion resistance of mudstone and siltstone is about 90 times larger than that of clayrock or redbeds, and about 16 times higher than that of sandstone.

What should be noted is, the actual situation of CBM preservation is much more complex than one-dimension gas diffusion, which also involves the CBM generation, migration and diffusion into other directions. With the subsidences and lifts of the strata, the diffusion coefficient of the caprocks will not stay constant as well. Moreover, interbedded patterns of caprocks and the fractures in rocks may also significantly affect the gas sealing capacities. In a word, to comprehensively describe the CBM preservation process, vast works are still needed.

4.3 Typical case in Xutuan Colliery

As mentioned in section 1, the unique tectonic movements of the Xutuan Colliery resulted in the different caprock groups in the southern and northern flanks of the III3 District (Figure 8), where the caprocks in the southern flank are mainly composed by redbeds and clayrock while in the northern flank are composed of ordinary coal measure strata (mudstone, siltstone, sandstone) and clayrock, leading to distinct CBM preservation conditions. And the mining activities of the colliery indicated different CBM occurrence in the southern and northern flanks of III3 District as well.

Figure 7 Correlations between CBM escape amount and escape time

Figure 8 Distribution of redbeds in III3 District, Xutuan Colliery, modified from Ref. [7]

At present, four working faces in the III3 district have been mined and one working face is just under recovery, of which, three working faces are on the southern flank (near or below the caprock mainly consisted by redbeds), while the others are on the northern flank (Figure 8). A calculation of the gas emission situations of the five working faces (Figure 9) indicates that the gas emission quantity seems to decrease with the increasing thickness of the redbeds, and the average gas emission quantities of the working faces near/below the redbeds caprocks are 6.22 m³/min (3₂33 working face), 8.09 m³/min (3₂35 working face) and 6.84 m³/min (3₂37 working face), which are much smaller than those working faces far from the deposition boundary of redbeds (3₂34 working face: 12.93 m³/min; 3₂36 working face: 14.61 m³/min). With the recovery moving from the coal seam capped by redbeds to the coal seam capped by Permian strata (like moving from 3₂35 working face to 3₂36 working face), the gas emission quantity shows significant increase tendency.

Figure 9 Correlation of Paleogene redbeds and CBM emission quantities

A statistic of the gas content in the Xutuan and Zhaoji Collieries confirms a low CBM content area (with an average CBM content of only 2.52 m³/t) in the coal seam that was covered by Paleogene redbeds [7], although the depth of the coal seam in this area is deeper than other coal seam covered by Permian strata, which also verifies the effects of caprocks’ sealing capacities on the CBM preservation.

5 Conclusions

1) SEM analyses indicate that due to the shallow burial depth, lighter compaction and cementation effects, Neogene clayrock and Paleogene redbeds contain a large number of micron-size pores, leading to highly developed pore structures and connectivities, thus weak CBM sealing capacities. While the pore development degree of Permian sandstone is less remarkable, and the pore structures of Permian mudstone and siltstone which are mainly composed of flaggy and schistose grains are much denser. Overall, the sealing capacities of old strata (Permian mudstone, siltstone and sandstone) which underwent a higher degree of compaction are better than the newly deposited strata (Neogene clayrock and Paleogene redbeds).

2) MIP based pore structure determinations indicate that the PSDs of clayrock and redbeds are mainly concentrated in transition pores (36.86% and 65.41%, respectively) and mesopores (51.31% and 25.22%, respectively). Whilst the PSD of sandstone is dominated by macropores (66.18%) and mesopores (14.71%). The PSDs of mudstone and siltstone are concentrated in micropores (31.25% and 50%, respectively) and transition pores (43.75% and 28.57%, respectively). With much higher porosities and appropriate PSDs, clayrock (porosity: 15.1245%), redbeds (porosity: 17.8660%) and sandstone (porosity: 6.1258%) are favorable for gas diffusion and penetration in caprocks, while mudstone (porosity:1.5231%) and siltstone (porosity: 1.4451%) favor the sealing of CBM.

3) The methane diffusion coefficient experiments demonstrate that the methane diffusion coefficients of clayrock and redbeds are about 6 times higher than that of sandstone, or about 90 times higher than that of mudstone and siltstone. Thus, as caprocks, mudstone and siltstone can seal the CBM best; while the sandstone is a little weaker; clayrock and redbeds have the weakest sealing capabilities of CBM.

4) Based on the Fick’s first law, one-dimensional CBM diffusion model is derived and solved. Results indicate that under the same thickness, the higher diffusion coefficient of the caprock leads to larger escape amount of CBM; the CBM sealing abilities of mudstone and siltstone are almost 100 times higher than that of clayrock or redbeds, and nearly 17 times higher than that of sandstone. With the increase of the caprock thickness, the gas escape amount decreases in a power function tendency. Case study from the III3 District, Xutuan Colliery verifies the effects of caprocks’ sealing capacities on the CBM preservation.

References

[1] TAO Ming-xin, SHI Bao-guang, LI Jin-ying, WANG Wan-chun, LI Xiao-bin, GAO Bo. Secondary biological coalbed gas in the Xinji area, Anhui province, China: Evidence from the geochemical features and secondary changes [J]. International Journal of Coal Geology, 2007, 71(2): 358–370.

[2] JIN Kan, CHENG Yuan-ping, WANG Wei, LIU Hai-bo, LIU Zheng-dong, ZHANG Hao. Evaluation of the remote lower protective seam mining for coal mine gas control: A typical case study from the Zhuxianzhuang Coal Mine, Huaibei Coalfield, China [J]. Journal of Natural Gas Science and Engineering, 2016, 33: 44–55.

[3] SIDDIQUI F I, PATHAN A G, UNVER B, ERTUN G. Sustainable lignite resource planning at Thar coalfield, Pakistan [J]. Journal of Central South University, 2018, 25(5): 1165–1172.

[4] WANG Shao-feng, LI Xi-bing, WANG De-ming. Mining- induced void distribution and application in the hydro- thermal investigation and control of an underground coal fire: A case study [J]. Process Safety and Environmental Protection, 2016, 102: 734–756.

[5] ZHANG Guang-chao, HE Fu-lian, LAI Yong-hui, JIA Hong-guo. Ground stability of underground gateroad with 1 km burial depth: A case study from Xingdong coal mine, China [J]. Journal of Central South University, 2018, 25(6): 1386–1398.

[6] ZHANG Hong, CUI Yong-jun, TAO Ming-xin, PENG Ge-lin, JIN Xiang-lan, LI Gui-hong. Evolution of the CBM reservoir-forming dynamic system with mixed secondary biogenic and thermogenic gases in the Huainan Coalfield, China [J]. Chinese Science Bulletin, 2005, 50(1): 30–39.

[7] JIN Kan, CHENG Yuan-ping, WANG Liang, DONG Jun, GUO Pin-kun, AN Feng-hua, JIANG Li-min. The effect of sedimentary redbeds on coalbed methane occurrence in the Xutuan and Zhaoji Coal Mines, Huaibei Coalfield, China [J]. International Journal of Coal Geology, 2015, 137: 111–123.

[8] LEUNG D Y C, CARAMANNA G, MAROTO-VALER M M. An overview of current status of carbon dioxide capture and storage technologies [J]. Renewable and Sustainable Energy Reviews, 2014, 39: 426–443.

[9] WU Yu-dong, JU Yi-wen, HOU Quan-lin, HU Sheng-biao, PAN Jie-nan, FAN Jun-jia. Comparison of coalbed gas generation between Huaibei-Huainan coalfields and Qinshui coal basin based on the tectono-thermal modeling [J]. Science China Earth Sciences, 2011, 54(7): 1069–1077.

[10] BAO Yuan, WEI Chong-tao, NEUPANE B. Generation and accumulation characteristics of mixed coalbed methane controlled by tectonic evolution in Liulin CBM field, eastern Ordos Basin, China [J]. Journal of Natural Gas Science and Engineering, 2016, 28: 262–270.

[11] MCINTOSH J C, WARWICK P D, MARTINI A M, OSBORN S G. Coupled hydrology and biogeochemistry of Paleocene–Eocene coal beds, northern Gulf of Mexico [J]. Geological Society of America Bulletin, 2010, 122(7, 8): 1248–1264.

[12] QIN Yong, TANG Xiu-yi, YE Jian-ping, JIAO Si-hong. Characteristics and origins of stable carbon isotope in coalbed methane of China [J]. Journal of China University of Mining and Technology, 2000, 29(2): 113–119. (in Chinese)

[13] KDZIOR S, KOTARBA M J, PKAA Z. Geology, spatial distribution of methane content and origin of coalbed gases in Upper Carboniferous (Upper Mississippian and Pennsylvanian) strata in the south-eastern part of the Upper Silesian Coal Basin, Poland [J]. International Journal of Coal Geology, 2013, 105: 24–35.

[14] SAGHAFI A, PINETOWN K. The role of interseam strata in the retention of CO₂ and CH₄ in a coal seam gas system [C]// 10th International Conference on Greenhouse Gas Control Technologies. Amsterdam, Netherlands, 2011, 4: 3117–3124. DOI:10.1016/j.egypro.2011.02.225.

[15] SU Xian-bo, LIN Xiao-ying, ZHAO Meng-jun, SONG Yan, LIU Shao-bo. The upper Paleozoic coalbed methane system in the Qinshui basin, China [J]. AAPG Bulletin, 2005, 89(1): 81–100.

[16] MENG Yan-jun, TANG Da-zhen, XU Hao, LI Chen, LI Ling, MENG Shang-zhi. Geological controls and coalbed methane production potential evaluation: A case study in Liulin area, eastern Ordos Basin, China [J]. Journal of Natural Gas Science and Engineering, 2014, 21: 95–111.

[17] PALMER I D, METCALFE R S, YEE D, RURI R. Coalbed methane reservoir valuation and exploitation [M]. Xuzhou: China University of Mining and Technology Press, 1996. (in Chinese)

[18] DU Shang, SHAN Xuan-long, YI Jian, LI Ji-yan. Controlling factors of high-quality volcanic reservoirs of Yingcheng Formation in the Songnan gas field [J]. Journal of Central South University, 2018, 25(4): 892–902.

[19] ZHAO Zi-long, ZHAO Jing-zhou, REN Hai-jiao, LI Jun, WU Wei-tao. Characterization and formation mechanisms of fractures and their significance to hydrocarbon accumulation: A case study of Lower Ordovician mid-assemblage Formations in central Ordos Basin, China [J]. Journal of Central South University, 2018, 25(11): 2766–2784.

[20] CHENG Yuan-ping, WANG Hai-feng, WANG Liang, ZHOU Hong-xing, LIU Hong-yong, LIU Hai-bo, WU Dong-mei, LI Wei. Theories and engineering applications on coal mine gas control [M]. Xuzhou: China University of Mining and Technology Press, 2010. (in Chinese)

[21] SAGHAFI A, JAVANMARD H, ROBERTS D. Parameters affecting coal seam gas escape through floor and roof strata [C]// Proceeding of the 10th Underground Coal Operators Conference. University of Wollongong and the Australian Institute of Mining and Metallurgy. 2010: 210–216.

[22] ZHANG Pei-he, JIN Xiu-liang, LIU Yu-hui, WANG Zheng-xi, LIU Na-na. Synthetical analysis on geological factors ccontrolling coalbed methane [J]. Procedia Earth and Planetary Science, 2011, 3: 144–153.

[23] BAO Yuan, WEI Chong-tao, WANG Chao-yong, WANG Guo-chang, LI Qing-guang. Geochemical characteristics and generation process of mixed biogenic and thermogenic coalbed methane in Luling coalfield, China [J]. Energy & Fuels, 2014, 28(7): 4392–4401.

[24] HONG Feng, SONG Yan, CHEN Zhen-hong, ZHAO Meng-jun, LIU Shao-bo, QIN Sheng-fei, FU Guo-you. Study on process and model of CBM dissipating [J]. Chinese Science Bulletin, 2005, 50(1): 134–139.

[25] LIU Guang-di, ZHAO Zhong-ying, SUN Ming-liang, LI Jian, HU Guo-yi, WANG Xiao-bo. New insights into natural gas diffusion coefficient in rocks [J]. Petroleum Exploration and Development, 2012, 39(5): 597–604.

[26] SAGHAFI A. Potential for ECBM and CO₂ storage in mixed gas Australian coals [J]. International Journal of Coal Geology, 2010, 82(3, 4): 240–251.

[27] NELSON J S, SIMMONS E C. Diffusion of methane and ethane through the reservoir cap rock: Implications for the timing and duration of catagenesis [J]. AAPG Bulletin, 1995, 79(7): 1064–1073.

[28] HUANG Zhi-long, HAO Shi-sheng. Study on sealing of gas concentration and diffusion in overlying gas reservoirs [J]. Acta Petrolei Sinica, 1996, 17(4): 36–41. (in Chinese)

(Edited by YANG Hua)

中文导读

基于实验与案例分析的盖层封盖能力对煤层气保存的影响

摘要：盖层的封盖能力对煤层气藏的保存起着至关重要的作用。为研究盖层的封盖能力，本文采集了五种不同岩性的盖层岩样（包括：新近系黏土岩、古近系红层、二叠系砂岩、二叠系泥岩和二叠系粉砂岩），综合采用显微结构观测、孔隙结构测量和扩散性能测定相结合的实验方法开展相关研究。研究结果表明：泥岩、粉砂岩由于其所具有的结构致密、孔隙率低、孔径分布以微孔和小孔为主以及孔喉比高等特征，是煤层气保存较为理想的盖层；甲烷在泥岩和粉砂岩中的扩散系数约为砂岩的6倍、粘土岩和红层的90倍。为进一步对煤层气通过盖层的逸散情况进行评价，本文构建了一维的煤层气扩散模型。建模分析结果显示：在相同盖层厚度条件下，泥岩、粉砂岩对煤层气的封盖能力是黏土岩和红层的近100倍、砂岩的近17倍。因此当煤层的盖层由黏土岩、红层或砂岩组成时，其下部煤层所遭受的瓦斯逸散作用将更为强烈。许疃煤矿III3采区的开采实践验证了上述结论，许疃煤矿的开采实践表明：与二叠系岩层(泥岩、砂岩、粉砂岩)相比古近系红层下方煤层的煤层气含量出现显著降低。

关键词：盖层；封盖能力；煤层气；保存

Foundation item: Project(2016YFC0801608) supported by the National Key Research and Development Plan, China; Project(51574148) supported by the National Natural Science Foundation of China

Received date: 2017-02-05; Accepted date: 2018-03-23

Corresponding author: LIANG Yun-tao, PhD, Professor; E-mail: liangyuntao@vip.sina.com; ORCID: 0000-0002-9253-5373; TIAN Fu-chao, PhD Candidate, Associate Researcher; E-mail: tianfuchao@cumt.edu.cn; ORCID: 0000-0002-7343- 1448