Liquid nitrogen (LN2) fracturing is a promising new technology for unconventional reservoir simulation because it can effectively solve problems related to low permeability, low brittleness, and water shortage. The present work conducted a series of permeability and strength property-related experiments to evaluate the effect of LN2 cooling on the permeability and mechanical characteristics of anisotropic shale. The main findings of the study are as follows: (1) The influence of the bedding direction on the permeability of anisotropic shale cannot be eliminated by LN2 cooling. LN2 cooling could effectively increase the initial natural damage and the pore space of anisotropic shale, possibly increasing the volume of reservoir stimulation and provide more channels for the seepage and migration of oil and gas.
Long Jiang1, Yuanfang Cheng1, Zhongying Han1, Qi Gao1, Chuanliang Yan1, Huaidong Wang2, Lipei Fu3
1School of Petroleum Engineering China University of Petroleum (East China) Qingdao China. 2CCCC Marine Construction & Development Co., LTD Tianjin China. 3School of Petroleum Engineering Changzhou University Changzhou China
Received: 23 January 2018 / Accepted: 10 June 2018
© The Author(s) 2018
(2) After LN2 cooling, the strength and brittleness of shale are obviously reduced, leading to the decrease in the ability of shale to resist deformation and failure, thereby helping to decrease the initiation pressure of reservoir stimulation. (3) The brittleness of shale will markedly increase during cryogenic fracturing, thus helping to form more complex fracture networks. Based on the present research, LN2 fracturing has obvious advantages compared with hydraulic fracturing in increasing the volume of reservoir stimulation. The results of this study are instructive for understanding the synergistic mechanism of LN2 fracturing and evaluating the effectiveness of reservoir simulation.
Shale gas, which has become an important unconventional natural gas resource all over the world, has exploitable reserves of approximately 207 × 1012m3, accounting for 32% of the total natural gas resources of the world (U. S. EIA 2011; Dong et al. 2012). The effective exploitation and utilization of shale gas are beneficial to increase the supply of clean energy, adjust the energy structure, and alleviate the shortage of oil and gas resources (Qiu et al. 2012; Dong et al. 2012).
However, the increase in shale gas production mainly depends on the exploitation of horizontal wells and the development of hydraulic fracturing technology (Middleton et al. 2015; U. S. EIA, 2018). At present, more than 90% of shale gas wells and 70% of oil wells are required to carry out large-scale hydraulic fracturing for effective exploitation (Brannon 2010; EIA, 2018), resulting in the explosive growth of hydraulic fracturing applications. Inspired by the global shale gas revolution, China is also successfully conducting shale gas exploration.
However, China’s shale gas reservoirs have the characteristics of a deep buried depth, large horizontal stress difference, low permeability and low brittleness index, making it difficult to form complex fracture networks after conventional hydraulic fracturing (Jiang et al. 2017). Moreover, as the shale gas reservoirs in China are mainly distributed in water-scarce regions, such as the Sichuan, Tarim, and Turpan-Hami Basins, and large-scale hydraulic fracturing requires the consumption of a large amount of water, hydraulic fracturing cannot be applied in these regions (Li et al. 2016b; Cheng et al. 2017).
Furthermore, the environmental problems caused by hydraulic fracturing, such as the consumption of freshwater resources (Arthur et al. 2009), groundwater contamination (Becklumb et al. 2015), earthquakes (Green et al. 2012), air pollution and clay expansion (Howarth et al. 2011), and storage and treatment of wastewater (Becklumb et al. 2015), has raised significant concerns from many scholars and governments. To meet the special requirements of shale gas exploration to reservoir according to the properties and geological environment, it is the focus of petroleum industry to improve the effect of volume fracturing and identify a substitute for the water-based fracturing fluid (Middleton et al. 2015; Wang et al. 2016).
Therefore, many new waterless fracturing technologies, such as liquefied petroleum gas (LNG) fracturing (Lestz et al. 2007), nitrogen foam fracturing (Gupta et al. 1998), liquid/supercritical CO2 fracturing (Gupta et al. 1998; Middleton et al. 2015), and cryogenic fracturing using LN2 (Mcdaniel et al. 1997; Grundmann et al. 1998; Cheng et al. 2017), have been considered by many scholars.
Because LN2 has an extremely low temperature and a high diffusion capacity at room temperature and atmospheric pressure, cryogenic fracturing using LN2 has been considered to be an ideal waterless fracturing technology for shale gas exploration (Li et al. 2016b; Cheng et al. 2017). When LN2 is in contact with rock, the temperature of the rock will drop sharply, thereby inducing a large number of micro-cracks in the interior and surface of the rock due to the non-uniform thermal stress. This approach can obviously improve the pore structure and reduce the bearing capacity of rock.
A few experimental studies on cryogenic fracturing using LN2 have been carried out. Mcdaniel et al. (1997) conducted in situ experiments using LN2 in 5 CBM wells, and the results varied: three wells showed increased production, one well showed equivalent production, and one tight sandstone well showed decreased production. Grundmann et al. (1998) exploited a Devonian shale well with LN2 and found that the initial production of the well was 8% higher than that of conventional fractured wells. Cha et al. (2014) and Alqatahni et al. (2016) also performed laboratory tests on cryogenic fracturing with LN2 and found that LN2 cooling could cause sandstone, concrete and shale to produce many interlaced micro-cracks, and decrease the initiation pressure by 40%.
Moreover, Cai et al. (2014, 2015) and Cheng et al. (2017) also noted that LN2 cooling can increase the pore connectivity and permeability of shale and coal and reduce their bearing capacity. Although considerable efforts have been made to develop the cryogenic fracturing using LN2, but these studies mainly focused on the change in rock micro-cracks after LN2 cooling, and only a few publications mentioned on the change in the damage characteristics and mechanical properties of rocks. In particular, they did not discuss in depth the influence of LN2 cooling on the seepage and mechanical characteristics of anisotropic shale with different bedding orientations.
As we know, shale has strong anisotropy due to the presence of foliation and schistose planes, which seriously affect the change in its pore structure and compressibility after LN2 cooling. Moreover, the influence mechanism of LN2 fracturing is still not completely understood at the bottom of the well. Considering the defects of previous studies, it is of great significance to study the influence of LN2 cooling on the seepage and mechanical characteristics of anisotropic shale.
Accordingly, in this study, cylindrical shale samples with different bedding angles (i.e., 0°, 30°, 60° and 90°) were used to carry out gas permeability, Brazilian splitting and triaxial compression tests. The influence of LN2 cooling on the permeability and mechanical characteristics of anisotropic shale with different bedding orientations were analyzed based on these experimental results. The study results are expected to provide a fundamental experimental basis for studying the synergistic mechanism of LN2 fracturing and the field application of cryogenic simulation.
Experimental items and testing procedure
Gas permeability test
Permeability, which is widely used to characterize the ability of soil and rock to conduct fluid, is also an important parameter for determining shale gas production. The permeability in rock depends mainly on its porosity, pore size and distribution, pore shape and the arrangement of the pores (AadnØy et al. 2011; Li et al. 2016a). Permeability increases with increasing porosity and grain size, whereas it decreases with the increase in rock compaction and cementation (AadnØy et al. 2011). In this study, gas permeability tests were conducted to measure the gas flow characteristics of anisotropic shale before and after LN2 cooling. Permeability is governed by Darcy’s Law, the typical calculation formula is given as follows:
where k is the permeability of the rock sample (m2); Q is the gas flow of the rock sample (m3/s); μ is the fluid viscosity (Pa.s); L is the sample length (m); A is the original cross-section area of rock sample (m2); p1 is the gas pressure at the inlet (Pa); and p2 is the gas pressure at the outlet (Pa).
Brazilian test is often used to measure the tensile strength of rock. In this study, shale has obvious anisotropic characteristics because of the presence of foliation and schistose planes; as a result, the traditional isotropic elastic theory is no longer applicable. In the Brazilian tests, the shale is assumed to be a transversely isotropic material. During the tests, it is worth noting that the shale samples should be inclined at four different angles with respect to the corresponding bedding direction, as shown in Fig. 1.
Fig. 1 Loading of shale disc-shaped sample
In the figure, θ is the inclination angle between the failure load and the normal direction of the bedding plane (0∘⩽θ⩽90∘). Based on the transversely isotropic elastic theory, Claesson et al. (2002) proposed that the analytical solution of tensile strength can be calculated by the Brazilian test results and the following expressions:
where σt is the tensile strength (MPa); P is the failure load (kN); D is the sample diameter (mm); t is the thickness of the Brazilian disc (mm); E is the elastic modulus in the transverse isotropic plane (GPa); E′ is the elastic modulus perpendicular to the transversely isotropic plane (GPa); G′ is the shear modulus in the plane perpendicular to the transversely isotropic plane (GPa) and is given by Eq. (4) because of the difficulty of measurement (Saint–Venant 1863); and ν′ is the Poisson’s ratio perpendicular to the transversely isotropic plane.
From Eqs. (2)-(4), five basic elastic constants were acquired via a series of uniaxial compression tests with different bedding shales according to these references (Amadei et al. 1996; Cho et al. 2012). Table 1 lists the mean value of the transversely isotropic basic elastic constants for the tested shale samples.
Table 1 Basic elastic constants of shale samples
Triaxial compression test
The use of the stress–strain curve of rock is an important method to reflect the mechanical properties of rock material and can fully reveal the deformation and failure characteristics of rock (Cheng et al. 2017; Yan et al. 2015, 2017). In this study, a series of triaxial compression tests were conducted to obtain the mechanical parameters of anisotropic shale; such tests are helpful for evaluating the mechanical characteristics of shale before and after LN2 cooling. Figure 2 shows a TAW-100 servo-controlled triaxial testing system and an assembled shale sample.
Fig. 2 TAW-100 servo-controlled triaxial testing system and assembled shale sample
During the tests, the sample was placed inside the heat-shrink rubber, and the fluid pressure (confining pressure) is applied to its surface. Axial and radial deformations of the specimens were most conveniently monitored by the deformation sensor. Four axial and lateral strains were recorded simultaneously by the axial and lateral strain gauges attached to the sample surface.
During the compression process, acoustic emission (AE) sensors are applied to monitor fracture initiation and extension in the shale specimens. Then, the stress–strain curves of these triaxial compression tests are recorded. Based on these test results, the influence of LN2 cooling on the energy storage limit and brittleness characteristics of anisotropic shale is also discussed in detail in the following sections.
The shale samples studied in the present experiments were obtained from the outcrop of the Silurian Longmaxi formation in Shizhu, Chongqing City, China. The mineral content of the shale was obtained by X-ray diffraction (XRD) analysis, as presented in Table 2. As a type of outcrop shale, natural shale has an initial water content of 15.4%. All shale samples were drilled from the same shale block along 4 different bedding directions. Figure 3 displays the specific core direction, and the dotted lines represent the bedding plane of the shale samples. The bedding angle (β = 0°, 30°, 60° and 90°) is defined as the angle between specimen axis and the layer orientation.
Fig. 3 Directional coring diagram for shale sample preparation
According to the International Society for Rock Mechanics standard (Bieniawski et al. 1979; Hawkes et al. 1978), the shale samples were processed into cylinders with 25 mm diameter and 50 mm height for the permeability and triaxial compression tests, and Brazilian discs with 25 mm diameter and 15 mm thickness were processed for the Brazilian tests. The length and diameter errors of all the specimens are within ± 0.5 mm and their parallelism at the end is within ± 0.02 mm after polishing. Water was not used during sample preparation.
Table 2 Mineral content of the shale specimen
To avoid differences among the samples, all drilling samples were screened by magnified observation using a microscope, and the samples without visual cracks were selected as the experimental specimens. In this study, Brazilian tensile specimens were marked as Bi-j, permeability and triaxial compression specimens were marked as Ci-j (i = 1, 2, 3, and 4 represent 0°, 30°, 60°, and 90°, respectively; j = 1, 2, and 3 represent original and untreated, dry and LN2-treated, and saturated and LN2-treated, respectively).
First, all required cooling samples (i.e., Bi-2, Bi-3, Ci-2 and Ci-3) were placed in the drying oven at a constant temperature of 60 °C until fully dried (if the temperature is too high, then the sample will crack.), and the compared samples (i.e., Bi-1 and Ci-1) were not dry-treated. Next, gas permeability tests were conducted on all the Ci-j samples, and the pressure of the inlet and the outlet of all tested samples were monitored until the inlet pressure remained constant for at least 30 min.
Second, all required water-saturation samples were vacuumed and immersed in distilled water until complete saturation. Next, all required cooling samples were completely immersed in LN2 for half an hour until fully cooled, and the compared samples were not cool-treated. The detailed cooling process can be seen in our previous reports (Cheng and Jiang et al. 2017). After the cooled samples recovered to room temperature, all the treated Ci-j samples were again subjected to the similar gas permeability test.
Finally, Brazilian and triaxial compression tests were conducted on all the Bi-j and Ci-j samples, respectively, following the International Society for Rock Mechanics standard (Bieniawski et al. 1979; Hawkes et al. 1978). During the Brazilian tests, all the Bi-j samples should be inclined at four different angles with respect to the corresponding bedding direction (see Fig. 1). During the triaxial compression tests, the confining pressure was applied to the Ci-j samples via a fluid pressure of 10 MPa (see Fig. 2). The loading rate for both tests was maintained at 0.05 mm/min.
Experimental results and analysis
Change in the permeability
Based on the results of previous studies, the better the pore connectivity of rock is, the higher value of permeability, i.e., the ability of rock to conduct fluid increases (AadnØy et al. 2011; Cheng et al. 2017). Figure 4 shows the change in the permeability of different bedding shales before and after LN2 cooling. First, it can be noticed that the permeability of shale always decreases gradually with increasing bedding angle in the three treatment methods. The maximum permeability always occurred at β = 0°, and the minimum value ways occurred at β = 90°. This result shows that the permeability of shale is influenced by the bedding direction because of the presence of foliation and schistose planes.
Fig. 4 Permeability of different bedding shales before and after LN2 cooling
The direction of the gas flow parallel to the bedding planes will have less impact on its flow capacity, which may be mainly attributed to the connectivity of anisotropic micro-cracks between the layers. Therefore, the permeability parallel to the bedding direction is significantly greater than those in other directions for the three treatment methods. This result reveals that the influence of bedding directions on the permeability of anisotropic shale cannot be eliminated by LN2 cooling. In addition, it can be observed explicitly that the permeability of all LN2-treated samples increases significantly compared with those of original and untreated ones.
Moreover, the increase in the permeability of the saturated and LN2-treated samples is greater than those of the dry and LN2-treated ones. For the bedding angles from 0° to 90°, the permeability of the dry and LN2-treated samples increased by 43.7, 21.5, 25.7 and 20.2%, respectively; the permeability of the saturated and LN2-treated samples increased by 66.3, 50.2, 53.7, and 41.3%, respectively. The experimental results reveal that the ability of shale to conduct fluid increases dramatically after LN2 cooling. The main reason may be that LN2 cooling induces a large number of micro-cracks and improves the pore size and connectivity in the shale. These micro-cracks provide the main paths for fluid flow, resulting in an increase in the diffusion and migration paths of fluid. Therefore, the permeability of the shale obviously increases after LN2 cooling.
Change in Brazilian tensile strength
The Brazilian tensile strength (BTS) of shale was obtained using Eqs. (2), in which shale was assumed to be transversely isotropic material. Figure 5 displays the change in the BTS of different bedding shales before and after LN2 cooling. It can be noted from the figure that the BTS of all shale samples always tends to increase with the increasing of bedding angle in three treatment methods. The minimum tensile strength always occurred at β = 0°, and the maximum value always occurred at β = 90°. Clearly, the tensile strength is affected by the bedding plane normal to the maximum tensile deformation direction. When the load direction was parallel to the bedding direction, the weak bedding planes of shale specimens maintain the maximum tensile stress normal to it, making it easier to generate micro-cracks.
Fig. 5 Brazilian tensile strength of different bedding shales before and after LN2 cooling
During the tests, it could also be observed that the cracks of the original shale samples propagate along the bedding directions at 0∘⩽β⩽60∘ and along the load direction at β = 90°. However, after LN2 cooling, the failure planes obviously exhibited symmetrical crossover along the bedding and load directions at β = 90°. This result shows that the effect of LN2 cooling on the strength of bedding weak planes is more serious than those of other regions. Comparing the changes in the BTS of the same bedding directions, it was found that the BTS of all bedding directions decreased to different degrees after LN2 cooling, and the BTS of all saturated and LN2-treated samples were lower than those of dry and LN2-treated ones. For the bedding angles from 0° to 90°, the BTS of the dry and LN2-treated samples reduced by 21.6, 12.5, 18.6, 20.2%, respectively, and the BTS of the saturated and LN2-treated samples decreased by 50.98, 44.6, 35.4, 37.1%, respectively.
Needless to say, the BTS of LN2 cooling samples is significantly affected by the steep temperature gradient and the degree of hydration. The main reason for the above phenomenon may be that LN2 cooling causes the high shrinkage of dried shale and the frost heaving of water-saturation shale, resulting in the expansion and coalescence of the initial natural micro-cracks and the generation of secondary micro-cracks. Moreover, in the process of water-saturation, when water intrudes into the shale pores, and interacts with the clay minerals, shale will undergo hydration expansion, resulting in the reduction of shale cohesion (Liang et al. 2015). Therefore, the internal structure of anisotropic shale is seriously damaged, resulting in the decrease in its tensile strength.