Abstract
In many engineering projects, it is critical to consider the acid erosion of rock. This study investigates dynamic changes in pore structure and damage mechanisms in sandstones subjected to acid erosion. Specimens with three grain sizes were immersed in acid solution and tested by the nuclear magnetic resonance technique. Changes in solution pH, specimen mass and porosity, T2 spectrum distribution, and area were analyzed. Damage mechanisms are discussed, and relationships between porosity and acid erosion damage variables are established. The results show that acid erosion has significant effects on pore structure and erosion damage in sandstone. With increasing soaking time, new micropores formed in sandstone, while existing micropores and mesopores gradually expanded into macropores, causing the T2 spectrum distributions to change greatly. The porosity, acid erosion damage, and T2 spectral areas of sandstones with different grain sizes all increased gradually. Under acid erosion, sandstones became gradually weakened, but the effects varied greatly according to grain size. Pore structure changes and acid erosion damage were greatest in coarse-grained sandstone, followed by medium- and fine-grained sandstone.
1. Introduction
Natural rocks have a large number of irregular pore structures that occur at multiple scales [1, 2]. Water is an active component of the geological environment that has an important influence on the safety of rock mass engineering [3, 4]. There is a variety of physical, mechanical, and chemical interactions that occur between engineering rock masses and surrounding water in deep geological environments [5]. These interactions are mainly caused by chemical reactions in which some of the active minerals in rocks are dissolved. The solutes migrate with the water out of the rock body, leaving caves and fissures that increase rock porosity [6] and cause erosion, thereby seriously affecting the safety of underground engineering projects [7].
In order to comprehensively study the macromechanical damage caused by chemical erosion of rocks, many experimental studies have been conducted. Lin et al. [2] investigated the effects of chemical erosion due to solution soaking on the pore structure and mechanical properties of sandstones by analyzing porosity, T2 spectrum distributions, and nuclear magnetic resonance images (NMRI). Han et al. [1, 8] used damage variables based on pore variation to quantitatively analyze sandstone damage caused by chemical erosion. Miao et al. [9] performed uniaxial and triaxial compression tests and splitting tests on granite samples after acid solution erosion at different pH values and flow rates. Deng et al. [10] studied the damage characteristics and mechanisms of sandstones subjected to acid erosion and repeated freeze-thaw cycles. Fang et al. [11] studied the combined effects of chemical erosion and freeze-thaw cycles on the compressive failure characteristics of yellow sandstones. Zhang et al. [12] studied pore structure changes in sandstones subjected to rapid freeze-thaw cycles and chemical erosion. Li et al. [13] studied the effects of acid solution type and soaking time on the physical and mechanical properties of sandstones. They measured changes in relative mass, deformation, and strength characteristics of sandstones under different acid-base erosion values. It can be seen that current research into the chemical erosion characteristics of rocks mainly focuses on the deterioration in their macromechanical properties caused by chemical reactions between solutions and minerals. However, the influence of rock grain size on microporous structures subjected to chemical erosion has received little attention.
Studies on rock pore structures and damage have adopted a variety of advanced testing techniques, such as mercury injection testing [14, 15], scanning electron microscopy (SEM) [16–18], computerized tomography (CT) [19–23], and NMR [24–27]. Each has its own advantages and disadvantages. Mercury intrusion testing can analyze the pore throats of pore structures, but the pores and throats cannot be identified separately [28, 29]. SEM can describe rock micropore structures but can only test a relatively small extent of the rock surface and so cannot provide overall pore structure information nor dynamically observe changes in pore structure. Use of CT can provide a complete three-dimensional view of rock samples but can only be used for real-time observations and cannot determine the characteristics of the rock microstructure [10] and is uneconomical [30]. NMR is a nondestructive testing technique. The dynamic evolution of a rock pore structure can be obtained by the measuring porosity and T2 spectrum distribution [2, 31]. It has been widely used to analyze pore structures and fluid states in rocks. Therefore, the present study used NMR technology to study dynamic processes and damage in rock pore structures.
In view of the above review, it is necessary to make an in-depth study of changes in the micropore structures of similar rocks with different grain sizes subjected to acid erosion. In order to study such dynamic processes, rock specimens were immersed in acid and then tested by the NMR technique. Changes in specimen mass, solution pH value, porosity, T2 spectrum distribution, and area at different time periods are comprehensively analyzed. The damage mechanisms occurring in sandstones with different grain sizes and subjected to acid erosion are discussed, and the relationship between sandstone porosity and acid erosion damage variables are established based on the test data.
2. Materials and Methods
2.1. Specimen Preparation
Sandstone test specimens were collected from the Zhiluo Formation in the Binchang mining area, Shaanxi Province, China. With reference to relevant standards [32], the samples were categorized according to grain composition as fine-, medium-, and coarse-grained sandstone. The grain sizes of fine-grained sandstone were 0.10–0.15 mm, those of medium-grained sandstone were 0.28–0.45 mm, and those of coarse-grained sandstone were 0.55–0.83 mm. The sampling depth was 705–720 m. All specimens were taken from intact sandstone and processed into cylindrical specimens with a height-to-diameter ratio of 1 : 1 and diameter of 25 ± 1 mm. In the tests, three specimens for each grain size were used, labeled groups B (N-1–N-3), C (N-4–N-6), and D (N-7–N-9), as shown in Figure 1. Before the tests, three groups of specimens were pretested by NMR, excluding specimens with large differences in initial porosity and T2 spectrum distributions. The main mineral compositions of the sandstones are shown in Table 1. Note that the results listed in Table 1 are the mean values of three specimens.
2.2. Test Instrument
The main equipment used in the tests was a NMR test system (MacroMR12-150H-1, Niumag Electric Technology Co., Ltd., Suzhou, China; Figure 2). The NMR tests required constant temperature conditions of 32.0 ± 0.02°C. The specific parameters are shown in Table 2.
2.3. Test Procedure
In a real geological environment, water-rock chemical interactions are slow, long-term processes. However, it is possible to alter the ionic concentration and pH of the test solution to accelerate the process [5, 7, 33]. In order to simplify the experiment, this work only investigates dynamic changes to the pore structures of sandstones under acid erosion with a solution of pH 2. The main test steps were as follows: (1) Preparation of specimens: All specimens were washed, dried at 105°C for 24 h, and cooled to room temperature in a drying dish. Their size and mass were measured and their appearance was observed. (2) Vacuum saturation: Specimens were placed in a vacuum saturation device, saturated for 4 h at a vacuum pressure of 0.1 MPa, and then immersed in distilled water for a further 24 h. (3) Preparation of acid solution. Three samples of identical 0.01 mol/L H2SO4 solution (pH 2) were prepared. (4) Testing: Specimen mass and porosity, solution pH, and NMR T2 spectra were determined. These tests involved no soaking (for 0 days). Then, the specimens were immersed vertically in H2SO4 solution and removed at 10 d intervals to repeat the above tests. (5) End of testing. Once the average porosity of a group of specimens became constant, all specimens were removed and the testing was complete. Figure 3 shows a complete flowchart of the test plan. For all tests, the same reaction vessels (1 L) and solution volumes (0.5 L) were used.
2.4. NMR Theory
The NMR technique is based on the interaction between the magnetic fields of hydrogen nuclei and the external magnetic field [7, 34]. Under the influence of static and alternating magnetic fields, the H protons in water-saturated specimens release energy that forms the NMR signal [35, 36]. Through these signal differences, the distribution of transverse relaxation time T2 can be obtained, which directly reflects the variation in the characteristics of the rock’s pore structure. T2 is the transverse relaxation time of fluid, which can be divided into three relaxation mechanisms in rock pores. The total transverse relaxation rate can be expressed as follows [37]:where is the transverse relaxation time; is the bulk fluid relaxation time; is the surface relaxation time; and is the diffusion relaxation time (all in ms). Because the diffusion relaxation and bulk fluid relaxation times are negligible compared with the surface relaxation time, the total relaxation time of the rock depends on surface relaxation [28, 38]. The surface relaxation is related to the specific surface area of rocks (the ratio of pore surface area to pore volume). The larger the specific surface area of the rock, the stronger the surface relaxation, and vice versa [39]. Therefore, the total transverse relaxation rate can also be expressed by the following formula [40, 41]:where S is the pore surface area (μm2); V is the pore volume (μm3); and is the transversal surface relaxivity (μm/ms). From equation (2), it can be seen that the T2 of rock is mainly determined by the ratio of and S/V. It can be further expressed as the relationship between T2 and pore radius:where is a dimensionless pore shape factor ( for cylindrical pores and for spherical pores) and r is the pore radius (μm). The transversal surface relaxivity () of rock has an important influence on the pore radius distribution, and different rocks have different values, generally ranging from 1 to 10 μm/ms [24]. In this paper, the middle value = 5 μm/ms is selected to study the pore radius distribution of sandstone. At the same time, according to the SEM images, the internal pores of sandstones with three kinds of grain sizes are mostly connected cylindrical pores, and is selected. Therefore, equation (3) can be simplified as
It can be seen that the NMR transverse relaxation time depends on the specific surface area of the rock [42]; that is, the T2 value is positively correlated with pore radius.
3. Test Results
3.1. Mass Change
The damage that acid erosion causes to rocks is mainly due to the infiltration of acid solution along pores and fissures and chemical reaction and dissolution with certain mineral crystals or framework grains. These processes increase the porosity of rock and make it more loose and fragile. The masses of sandstone specimens with different grain sizes after soaking in acid solution for 60 d were measured (Figure 4). It can be seen that with increasing soaking time, the masses of the specimens in each group increased to different degrees, but under the action of erosion, the masses of coarse- and medium-grained sandstone specimens decreased. This is because, under acid erosion, the internal pores expand continuously, leading to an increase in the water content of the specimens in the saturated state and an increase in their mass. However, with increasing soaking time, some mineral crystals or framework grains were dissolved and corroded to varying degrees, resulting in the shedding of framework grains from the surfaces of the rock specimens, which decreased their mass.
The relative mass change rate can indirectly describe the effect of acid erosion on a rock pore structure. Based on the difference between a specimen’s initial mass and that at time t, the rate of relative mass change Δm of specimens at different times can be calculated according to the following formula:where m0 and mt are the masses of specimens at the initial time and time t. Figure 5 shows the relationship between the rate of relative mass change and soaking time. It can be seen that there are different relationships according to the sandstone grain size. The test results show that(1)The rate of relative mass change in fine-grained sandstone increased slowly at first and then tended to be stable with increasing soaking time. The results show that the water absorption effect of the rock was greater than the effect of acid erosion in the initial stage of soaking. With increasing soaking time, acid erosion progressed and the micropores in the rock increased. However, at this time, the water absorption effect was still dominant and the masses of the rock specimens increased, so the rate of relative mass change increased slowly. When soaked for about 60 d, the water absorption effect of the rocks was basically the same as that of acid erosion. That is to say, after increases in rock pore numbers, the mass of water absorbed increased but was basically equal to the mass lost due to acid erosion, such that there was no obvious change in mass. At this stage, the relative mass change rate of the rocks only increased by 0.04%.(2)The rates of relative mass change in medium- and coarse-grained sandstones increased at first and then decreased with increases in soaking time. This was because, in the initial stage of soaking, the water absorption effect of rocks was greater than that of acid erosion, but with increasing soaking time, the porosity and specific surface area of the rock increased, and the rate of chemical reaction increased. At this time, acid erosion appeared to be the main effect, and rock mineral components were hydrolyzed and dissolved. Some framework grains were mechanically detached and the masses of the specimens decreased. Compared with medium-grained sandstone, the mass of coarse-grained sandstone decreased more obviously. After soaking for 60 d, the average mass of coarse-grained sandstone decreased to 27.20 g, a relative mass change rate of −3.03%.
3.2. Specimen Appearance and Changes in Grain Precipitation
The effect of acid erosion on the pore structures of the specimens changed their appearance to a certain extent. Figure 6 shows photographs of the specimens of different grain sizes after soaking for 60 d. It can be seen that the fine-grained sandstone appeared to be the most complete. Macropores were hardly observed, and a small denudation area was found at the bottom of the specimen. The medium-grained sandstone had a relatively complete appearance, with a honeycomb-like local surface and a small number of macropores, while the denudation area at the bottom was more obvious than in fine-grained sandstone. The coarse-grained sandstone appeared incomplete, the surface was honeycomb-like, there were many macropores, and the denudation area expanded to the periphery.
(a)
(b)
(c)
Changes in the amount of grains precipitated in the solution used to soak the specimens can also reflect changes in the pore structures of sandstones to a certain extent. In the process of acid erosion, precipitated grains were mainly formed by dissolution and mechanical shedding of framework grains from the rocks. Figure 7 shows the changes in precipitated grains in the solution after soaking for 60 d. It can be seen that there were basically no framework grains in the group B solution, free grains were few and scattered, and the solution was relatively clear. Free grains and a small amount of framework grains appeared in the group C solution. However, a large number of free grains and framework grains appeared in the group D solution, making it turbid and indicating that the reaction between coarse-grained sandstone and the acid solution was strong, and changes to the pore structure were obvious.
(a)
(b)
(c)
3.3. Solution pH
The saturated specimens were vertically immersed in H2SO4 solution and its pH was measured every 10 d (Figure 8). With increasing soaking time, the pH of three groups of solutions tended to be neutral, mainly because the water-rock chemical reaction consumed the H+ ions in the solution. In addition, the rates of pH increase differed between the three solutions at different stages. This indicates that the rates of water-rock chemical reaction differed between the three groups at different stages.
3.4. Porosity Change
Porosity change is of great significance in the study of rock pore structures subjected to acid erosion [43]. Figure 9 shows that the porosity of the specimens gradually increased with soaking time. However, the rates of change differed according to grain size. After 60 d of soaking, the porosities of fine-grained, medium-grained, and coarse-grained sandstone increased by 163.3%, 225.0%, and 271.0%, respectively. The rate of change was greatest in coarse-grained sandstone, and the porosity increased rapidly in the early stages of soaking. The main reason is that the initial pore-throat size of the coarse-grained sandstone was large, such that the acid solution could easily enter the interior and cause high initial dissolution. It is worth noting that the average porosity of coarse-grained sandstone was 15.55% after soaking for 60 d. Hence, porosity only increased by 0.02% compared with that at 50 d, which is negligible. This indicates that the water-rock chemical reaction had basically reached an equilibrium state. Comparing the groups of specimens, there is a positive correlation between the initial pore-throat size and increment of porosity. That is to say, under the same conditions, without considering the anisotropy of rock framework grains, a greater initial pore-throat size allows greater acid erosion damage and a more rapid increase in porosity.
(a)
(b)
(c)
3.5. NMR T2 Spectrum Distribution
Change in an NMR T2 spectrum can directly reflect change in the pore structure of rock [36, 41, 43]. According to principles of NMR, the pore radius is proportional to the value of T2, where a higher T2 indicates a larger pore radius. The peak value and peak area of the T2 spectrum distribution represent the concentration degree and quantity of different-size pores. Therefore, the effect of acid erosion on the pore structure of sandstones with different grain sizes can be studied by T2 spectrum distributions.
However, there is no unified classification standard for the pore radius of sandstones, and sandstones from different areas have different classification standards. According to the pore classification standard proposed by Ondrášik and Kopecký [44] and the pore structure characteristics of the specimens, the pores were divided into three types: micropores (≤100 μm), mesopores (100–1000 μm), and macropores (≥1000 μm). Based on this, the critical pore radius values of these pore types were substituted into equation (4), and the T2 cutoff time of the three pore types were calculated to be 10 ms and 100 ms. The T2 spectrum distribution curves of typical specimens from each group are analyzed as follows.
Figure 10 shows the T2 spectrum distribution curves of sandstones with different grain sizes under acid erosion. The T2 spectrum for each grain size of sandstone is composed of three peaks, each corresponding to a pore-size type. From left to right, the first signal peak is the strongest while the second and third ones are relatively weak. These peaks correspond to three pore types: micropores, mesopores, and macropores. At the same time, the peaks also indicate the degree of damage at this time. According to the T2 spectrum distributions, the damage can be defined as three types: microscopic, mesoscopic, and macroscopic. When the amount of macroscopic damage accumulates to a certain extent, the rock specimens will be destroyed.
(a)
(b)
(c)
Comparing Figures 10(a)–10(c), it can be seen that at a given soaking time, the signal amplitude detected in coarse-grained sandstone was the highest, followed by those of medium- and fine-grained sandstone. This indicates that the coarse-grained sandstone had the greatest number of pores. The maximum relaxation time of coarse-grained sandstone was 541.58–821.43 ms, that of medium-grained sandstone was 471.37–766.34 ms, and that of fine-grained sandstone was 333.13–439.76 ms. The higher relaxation time of coarse-grained sandstone indicates that it had larger maximum pore sizes than the other two sandstone types. In addition, from the T2 spectrum distribution, it can be seen that the order of influence of acid solution on sandstone pore structure was coarse-grained > medium-grained > fine-grained. It is worth noting that after 60 d of acid soaking, the third peak of the coarse-grained sandstone had an amplitude of 48.3 and its macroscopic damage was greater. The average porosity of coarse-grained sandstone obtained above was basically unchanged, indicating that the water-rock chemical reaction reached equilibrium. Under further soaking, coarse-grained sandstone should be destroyed first.
We can also see that with increasing soaking time, the T2 spectrum peaks of sandstone shifted left or right to various degrees and also fluctuated up and down. This is because, in the process of acid soaking, the mineral components of the rocks were hydrolyzed and dissolved. New micropores were continuously produced and, at the same time, some of the original micropores and mesopores gradually expanded into macropores. This shows that the pore structure changes occurring in sandstone due to acid erosion are a dynamic process.
3.6. T2 Spectrum Area Analysis
The T2 spectrum area is proportional to the amount of fluid contained in the rock; that is, it is related to porosity [42]. Therefore, under acid erosion, changes in the T2 spectrum areas of rock specimens reflect dynamic changes to the pores. The rate of change in the spectrum area in each stage also reflects the rate of damage to the rock specimens. Table 3 shows the NMR T2 spectrum area of sandstones and the rate of change in the spectrum area at each stage with soaking time.
As can be seen from Table 3, the T2 spectrum areas increased with soaking time, but the range of increase differed for each specimen. At a given soaking time, the rates of change in spectrum areas differed markedly according to grain size. At the same time, with increasing soaking time, the rates of change in the spectrum areas of sandstones with the same grain sizes also differed. This indicates that the rate of damage to sandstones with the same grain size differed according to soaking time. Hence, the damage rate of fine-grained sandstone changed from relatively slow to medium and then to slow. This means that in the initial stage of soaking, the initial porosity of fine-grained sandstone was low and the water-rock chemical reaction was slow. As soaking time increased, the pore sizes and quantity increased and the water-rock chemical reactions accelerated. However, with further erosion, due to a gradual decrease in the amounts of framework grains and cements involved in the chemical reactions, the seepage paths in the sandstone elongated. The cements, cations, and water films produced by chemical reactions block micropores inside rock, resulting in slowing down of water-rock chemical reactions and decrease in the rate of change in spectrum area. The rates of damage to medium- and coarse-grained sandstones changed from relatively fast to medium and then to slow, and from fast to medium and then to slow, respectively. The main reason is that the initial porosity of coarse- and medium-grained sandstone was higher than that of fine-grained sandstone, the initial water-rock chemical reaction was faster, and rate of change in spectrum area was greater. With increasing soaking time, the seepage paths in coarse- and medium-grained sandstone became longer; however, the main reason for the gradual slowdown in the water-rock chemical reactions was that the amounts of participating framework grains and cements gradually decreased.
In addition, the initial spectrum areas of the three specimens in each group differed, as did their increases with soaking. However, there was no obvious overall relationship. This is mainly due to the fact that rocks are complex porous materials with individual differences and poor homogeneity.
4. Discussion
4.1. Acid Erosion Damage
When studying the damage to rocks caused by acid erosion, it is critical to select appropriate variables. Changes in rock micropore structures are the fundamental cause of changes in rock mechanical properties [45, 46]. Therefore, porosity can be used as a variable to reflect the degree of acid erosion damage in sandstone. According to analysis of the dynamic change in sandstone porosity under acid erosion, an acid erosion damage variable, S, was established as follows:where φ0 is the porosity before erosion and φt is the porosity after acid erosion for time t. According to equation (6), S was calculated for all sandstone samples (Figure 11). It can be seen that S increased to varying degrees with acid erosion. The degree of damage to coarse-grained sandstone was always greater than that of medium-grained sandstone and was least in fine-grained sandstone.
4.2. Analysis of the Damage Mechanism
The micropore structure of rock is closely related to its macromechanical properties. Acid erosion changes the micropore structure, leading to deterioration in the macromechanical properties to varying degrees. It can be said that the essence of acid erosion is a process of changing the micropore structure and composition of rocks [45]. Under acid erosion, the micropore structures of the sandstone samples changed in two main aspects. On the one hand, the acid solution contained a large amount of H+ ions, which can easily enter the interior of rock specimens along pores and fissures of various scales on the sandstone surface. There, they react with minerals such as calcite and destroy the cementation between rock grains, making the interior loose and more porous. On the other hand, some minerals are readily soluble in aqueous solutions, resulting in increased porosity and softening of the interior of the samples. Figure 12 shows the water-rock chemical reaction model of sandstone.
Sandstone immersed in acid solution undergoes a series of complex chemical reactions, the main ones being the following:
As can be seen from Formulas (7)–(12), the chemical reactions in rocks differ according to the mineral composition and cause different types of chemical erosion damage. In this study, quartz was the main mineral in the sandstone samples, accounting for a large proportion of the total mineral contents. However, its main component is SiO2, which is relatively stable in acid environments and does not react. Feldspar, muscovite, kaolinite, and other framework grains will undergo various degrees of chemical reactions in acid environments, resulting in changes to the rock pore structure that make the rock softer. In addition, the cement content of sandstone is low, but because the strength of the minerals in the cements is often less than those of framework grains, chemical reactions can destroy the cementation between framework grains and other grains, such that the internal grains of the rock become looser. Therefore, chemical reactions involving cements have a greater effect on the micropore structure of rocks than other types of reactions. Based on the above analysis and the mineral compositions of the sandstones, it can be found that the changes in pore structure and amount of erosion damage are positively correlated with the contents of framework grains and cements that participate in chemical reactions.
In summary, due to chemical reactions and dissolution in acid, the micropore structure of sandstone changed, the porosity increased, and acid erosion damage increased. This conclusion is easily verified by the NMR T2 spectrum. The pore structure changed more and the acid erosion damage was greater in coarse-grained sandstone than in other types, which is attributed to the contents of framework grains and cements that can participate in chemical reactions and the initial rock pore-throat size.
5. Conclusions
In this work, through a series of laboratory tests, dynamic pore structure changes and the mechanism of damage in sandstones with different grain sizes subjected to acid erosion were studied. The following conclusions can be drawn:(1)With increasing soaking time, the porosity of sandstone increased gradually. However, the increment of porosity differed according to grain size. Comparing the three groups of specimens, a positive correlation was found between the initial pore-throat size and the increment of porosity.(2)After acid soaking, the pore structures and T2 spectrum distributions of the three sandstone types changed obviously. According to the T2 spectrum distribution curves, the order of influence of acid solution on the pore structures of sandstone was coarse-grained > medium-grained > fine-grained.(3)The rates of damage to sandstone specimens with different grain sizes differed according to the soaking stage. The rate of change in the T2 spectrum area at each stage reflected the rate of damage. Therefore, according to the rate of change in the T2 spectrum area, the damage rates of fine-grained, medium-grained, and coarse-grained sandstone were (1) relatively slow to medium and then to slow, (2) relatively fast to medium and then to slow, and (3) fast to medium and then to slow, respectively.(4)The degree of acid erosion damage to sandstone was ranked coarse‐grained > medium-grained > fine-grained. The contents of framework grains and cements that participate in chemical reactions, and the initial pore-throat size of rock, were larger in coarse-grained sandstone. Hence, it had the strongest water-rock chemical reactions, greatest changes in micropore structure, and greatest amount of erosion damage.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This study was supported by funds from the Open Project Fund of Key Laboratory of Coal Resources Exploration and Comprehensive Utilization, Ministry of Land and Resources (ZZ2016-1).