Impact of Pyrite Oxidation on the Pore-Structure Characteristics of Shale Reservoir Rocks under the Interaction of Fracturing Fluid.

Hydraulic fracturing combined with horizontal drilling is widely used to develop shale gas resources, and huge amounts of fracturing fluid are injected into shale reservoirs. However, the fracturing fluid is ineluctably retained in reservoir rocks after fracturing, resulting in the alteration of shale pore systems and further affecting the hydrocarbons production efficiency. In this work, two types of shales with different pyrite contents, namely, pyrite rich (PR, Niutitang Formation) and pyrite poor (PP, Xiamaling Formation), were emphasized to illustrate the effect of pyrite oxidation on pore structure after fracturing operation. Slickwater fracturing fluid was used to treat the shale samples for a period of 3 days, under the condition of 100 °C and 50 MPa. The field emission scanning electron microscopy (FE-SEM) and X-ray diffraction (XRD) were utilized to determine the surface morphology and mineral composition. The low-temperature N2 adsorption was performed to quantify the pore structure. The results showed that the pyrite oxidation induced the dissolution of both the pyrite and calcite and generated many dissolution pores for the pyrite-rich shale after slickwater treatment. The mineral dissolution led to an increase in the number of mesopores, enlarged the total specific surface area (TSSA) and total pore volume (TPV), and strengthened the pore-structure complexity. On the other hand, the pyrite-poor shale only experienced clay swelling after slickwater treatment. Its pore surface roughness and pore-structure complexity degraded with the loss of nanopores and the reductions in TSSA and TPV. The results of this study enhance the understanding of the impact of pyrite oxidation on the pore structure and provide new insight into the optimization of fracturing operation conditions based on shale's mineral composition characteristics.

Introduction

As one of the unconventional natural gas resources, shale gas is altering the global energy consumption structure. In the past decades, shale gas exploration and development were mainly concentrated in the United States, China, Canada, and Argentina, with production levels in 2018 being 6.072 × 1011, 1.09 × 1010, 5.3 × 109, and 4.3 × 109 m3, respectively.[1] The rapid increase in shale gas production around the world is attributed to the development of technologies in hydraulic fracturing and horizontal drilling. However, shale gas production decreases sharply after initial recovery from the hydraulically fractured wells.[2,3] One possible reason for the decrease in shale gas production is that the chemical reactions between minerals and fracturing fluid can alter the pore-structure characteristics of shale reservoirs.[4,5] Hydraulic fracturing introduces huge amounts of water-based fluid mixed with quartz sands and chemicals (7000–50 000 m3 per well) to generate complex fracture networks and create fluid flow pathways.[6,7] After the fracturing operation is complete, only 20–40% of the injected fluid flows back and more than 60% remains in the subsurface.[8] The retained fracturing fluid can damage the pore throats and fractures of the shale formation.[9] Therefore, efforts to increase shale gas production improve the importance of interactions between shale and injected fracturing fluid.

Shale as a kind of typical clay rock is primarily composed of clay minerals, brittle minerals (quartz, carbonates, feldspar, and chert), and easily oxidized components such as organic matter and pyrite.[10] Many researchers have revealed that the fracturing fluid chemically reacts with calcite, pyrite, and feldspar.[5,11−13] These chemical reactions can cause mineral dissolution and precipitation, alter rock properties, such as porosity and permeability, and further affect gas flow pathways within the shale matrix.[14] Thus, the relationship between geochemical shale-fracturing fluid interactions and the decline in shale gas production should be given more consideration. Herz-Thyhsen investigated the geochemical behavior of Al-bearing and Si-bearing minerals rock during simulated hydraulic fracturing treatment and found that the calcite dissolution increased the porosity of calcareous mudstone by 37%.[2] Lu also presented that the calcite and dolomite dissolution created up to 28% surface porosity of shale after reacting with fracturing fluid.[12] Xiong reported that large euhedral barite (10–50 μm) precipitated on the shale fracture surface during the simulated injection period.[15]

As a common mineral in organic-rich shale, pyrite not only has an important effect on the shale gas enrichment[16] but also has the characteristics of easy oxidation under the oxygen environment.[17] It has been illustrated that pyrite has positive correlations with liquid hydrocarbon and total organic carbon (TOC), and it can promote the pyrolysis of organic matter and catalyze hydrocarbon generation.[18] Chen demonstrated that the pores within the pyrite framboids, which were as high as 5.66%, had a favorable contribution to the pore system of shale reservoir, and these pores could reserve free gas and promote gas enrichment.[19] In addition, the presence of pyrite in shale reservoirs can accelerate the oxidation of shale. A previous study indicated that most fracturing fluids contained dissolved oxygen (9 mg/L) and some oxidants, which were used as glue-breaking agents.[20] Although various oxygen scavengers and Fe-controlling agents are added to fracturing fluids to prevent the oxidative dissolution of pyrite, fracturing flow-back fluid normally contains large amounts of Fe in solution, indicating that these additives do not prevent pyrite oxidation effectively in reservoir conditions.[5] Xu observed that the dissolved oxygen in fracturing fluid can cause the pyrite oxidation and dissolution in shale through water imbibition experiments under oxic and degassed conditions.[21] The oxidation of pyrite has been considered one of the key factors that influence the shale pore structure.[5,11,22] Jew found that the Fe(II)-bearing phases could change to Fe(III)-bearing precipitates when the pH of fracturing fluid was above 3.25 in the carbonate-poor shale.[5] However, You reported that the oxidation of pyrite could produce dissolution pores and cracks, and the oxidation sensitivity was controlled by the pyrite content, dissolved oxygen content, and reservoir conditions.[23] Moreover, Li revealed that the pyrite oxidation zone was >0.5 cm into the shale matrix during hydraulic fracturing via synchrotron X-ray-based techniques, indicating that the pyrite oxidation occurred not only close to microfractures but also in a wider range of shale reservoir matrix.[24] Owing to the complex mineral composition in organic-rich shale, the mechanism of the pore structure change induced by pyrite oxidation is still not well-understood. Although most of the previous studies have demonstrated that the dissolution of pyrite has a significant effect on shale pore structure, little is known about the positive or negative effect of pyrite oxidation on pore-structure modification in shale reservoirs during hydraulic fracturing.

As a result, we performed laboratory experiments to confirm the positive or negative effect of pyrite oxidation on the pore-structure characteristics of shale during hydraulic fracturing. In this research, two shale samples with distinct pyrite content obtained from Niutitang Formation (pyrite rich) and the Xiamaling Formation (pyrite poor) were used for the fluid–shale interaction experiments. The surface morphology, mineral composition, and pore-structure characteristics of shale samples before and after fracturing fluid treatment were analyzed. The impact of pyrite oxidation on shale reservoir rocks was discussed, and the chemical influential mechanism of fluid–shale interaction on the shale pore structure was revealed. The results of this work will be helpful in optimizing hydraulic fracturing operations for reservoirs with different mineralogical compositions and developing shale gas production.

Material and Methods

Shale Samples

Experimental shale samples used in this study were outcrops and obtained from Lower Cambrian Niutitang Formation (pyrite rich, labeled as PR) in the Youyang area of southeast Chongqing city and Mesoproterozoic Xiamaling Formation (pyrite poor, labeled as PP) in Chicheng area of northeast Hebei Province in China. These samples were selected due to their significant difference in pyrite content, and the basic information of the samples is shown in Table . Based on the XRD results, the major minerals for PR shale are quartz (65%), clay mineral (14%), and pyrite (8%), while quarte (82%) and clay mineral (14%) are the dominant minerals for the PP shale. In addition to the different mineral compositions, the TOC content of the PR shale is 8.4 wt %, higher than that of 1.1 wt % for the PP shale.

Table 1

Basic Information of Shale Samples Selected in This Study

		mineral composition (%)
sample ID	TOC (wt %)	quartz	pyrite	dolomite	calcite	feldspar	clay mineral
PR	8.4	65	8	1	1	11	14
PP	1.1	82				4	14

Prior to the experiment, the outcrop block shale samples were cylindrically drilled to 25 mm in diameter and 50 mm in length. They were used for XRD and low-temperature nitrogen (N2) adsorption tests before and after shale-fracturing fluid interaction. The thin slices of 2–3 cm3 were prepared for FE-SEM measurement before and after the reaction. Considering that slickwater has been commonly used in shale gas production,[25] slickwater was also selected in this study. The physical property and chemical additives of the slickwater were reported in our previous study.[26]

Experiments and Methodology

To simulate the short-term influence of fracturing fluid on the pore structure of shale under the reservoir condition, the experiments were performed at reservoir temperature (100 °C) and pressure (50 MPa) for three days using a high-temperature and -pressure reaction device. The selection of temperature and pressure was ascribed to the burial depth of the PR and PP shale (2500–4000 m, the geothermal gradient is 20–25 °C/km) and the average pressure gradient of 0.015 MPa/m.[27−30]

The schematic diagram and main components of the simulation instrument have been described in detail in our previous study.[13] The experimental procedures were given as follows. The prepared shale samples were placed into the reactor. Then, fracturing fluid was injected and kept at the pressure of 50 MPa. The fluid/rock mass ratio was approximately 10:1. Finally, a heating system was performed. During the experiment, 10 mL of each solution was collected (24, 48, and 72 h), and the pH of the solution was monitored. After the reaction, the shale samples were cleaned using deionized water and heated to dry at 50 °C for 12 h in an oven to minimize the influence of free water and weakly bounded water in shale pores.[31] Then, they were collected for analyzing the mineralogy and pore-structure characteristics.

To better observe the mineral precipitation and dissolution at the water–rock interface, a Merlin Compact field emission scanning electron microscope (FE-SEM) with energy-dispersive spectroscopy (EDS) was applied to the image. Before the experiment, the thin slices of shale samples were ion milled to reduce the artifacts and roughness effect on the observation by an argon ion mill instrument (PECS II 685 C).

X-ray diffraction (XRD) analyses (Rigaku D/Max-III B X-ray diffractometer) were performed to measure the mineral composition of unreacted and reacted shale samples. The cylindrical shale samples were crushed and passed through a 200 mesh sieve. The measurement was performed at 40 mA and 40 kV with Cu Kα radiation and scanned from 3 to 70° with a scanning rate of 2.0°/min.

To quantitatively analyze the pore evolution during hydraulic fracturing, a low-temperature N2 adsorption test was implemented at the ASAP 2020 HD88 surface area analyzer. The unreacted and reacted shale samples were crushed to 40–60 mesh, and the measurement process was identical to the study by Li which provided a detailed description.[32] The Brunauer–Emmett–Teller (BET) and Barrette–Joynere–Halenda (BJH) theories were used to calculate the specific surface area, pore volume, and pore size distribution.[33,34]

Based on the results of low-temperature N2 adsorption, the fractal dimension (D) of the pore structure was calculated. The Frenkel–Halsey–Hill (FHH) model was used to calculate the value of D by applying eqs and 2.[35,36]where P0 and P are the saturated vapor pressure and adsorption equilibrium pressure, respectively, MPa; V is the N2 adsorption volume at P, cm3/g; A is the slope of the linear fitting curve; and C is the fitting constant. D, which ranges from 2 to 3, can be calculated based on eq .

Results and Discussion

FE-SEM Imaging

Based on the FE-SEM, the variations in pore morphology and mineral composition on the shale surface were directly analyzed (Figures and 2). The pore type in the original PR shale mainly consisted of interparticle (interP) pores developed among the edge of brittle mineral particles and intraparticle (intraP) pores within rigid particles (Figure a,b). In addition, the shrinkage organic matter pores (OMPs) (Figure a) and clay mineral intercrystalline pores (Figure b) were developed in the original PR shale. However, Figure c,d displays that the original PP shale contains many clay mineral intercrystalline pores and interP pores develop between clay and brittle mineral particles, as well as the intraP pores related to brittle mineral dissolution.

Figure 1

FE-SEM images of different pore types in the original PR (a, b) and PP (c, d) shale.

Figure 2

FE-SEM images of the fracturing fluid-reacted PR (a, b) and PP (c, d) shale.

After the interaction experiments of slickwater fracturing fluid and shales, the changes in pore morphology and mineral composition on the shale surface are illustrated in Figure . Although the FE-SEM observation positions of the reacted shale are not the same as those of the unreacted shale, the qualitative analysis of the shale surface with similar morphology can still provide evidence of the changes on the surface of the matrix induced by water–rock interactions.[37] In the reacted PR shale, the original smooth surface of pyrite becomes rough, and pyrite shows obvious signs of erosion, which indicate that pyrite is dissolved and forms many dissolution pores within it. Meanwhile, clay swelling is also observed on its surface (Figure a,b). However, the original tight and flat shale surface of the PP shale matrix (Figure c,d) become expanded and looser after the slickwater treatment (Figure c,d), and the clay disintegration extends along the original clay mineral intercrystalline pores, which is closely related to the hydration expansion of the clay minerals.[38] The FE-SEM results indicate that the mineral composition has a significant effect on the pore structure of shale, and the reasons for the different variations of the two shale samples will be discussed later in this paper.

XRD Analysis

The changes in shale mineral composition before and after reaction with slickwater fracturing fluid are listed in Table . The analysis results suggested that the mineral compositions of the unreacted and reacted shale samples were different. In the PR shale, the contents of pyrite and calcite decreased by 25 and 100%, respectively (Table ). Combined with the FE-SEM observation results of the PR shale, the content variations of pyrite and calcite indicated that the dissolution of these minerals might have occurred during the experiment. Moreover, the content of dolomite and feldspar in the PR shale did not change, and the slight increase in quartz and clay minerals might be attributed to partial mineral dissolution. However, the reacted PP shale showed similar mineral composition to the unreacted sample (Table ). A previous study illustrated that the interaction between water-based fracturing fluid and shale could cause clay swelling,[39] which was in accordance with our results that the content of clay mineral decreased in the reacted PP shale. In addition, FE-SEM results of the reacted PP shale confirmed that the clay swelling occurred during the experiment.

Table 2

Mineralogical Compositions of the Unreacted and Reacted Shale Samples (%)

samples	state	quartz	pyrite	dolomite	calcite	feldspar	clay mineral
PR	unreacted	65	8	1	1	11	14
	reacted	66	6	1	0	11	16
PP	unreacted	82				4	14
	reacted	83				5	12

Owing to the pH variation of the reaction fluid reflecting the degree of fluid–shale interaction, the pH of fracturing fluid discharged in different time periods (24, 48, and 72 h) was determined during the experiment (Figure ). With the increase in reaction time, the pH of the aqueous solutions of the reacted PR and PP shale showed different trends. The pH for the PR shale remained acidic throughout the experiment, with a rapid decrease from 5.23 to 2.76 after 24 h and a gradual increase from 2.76 to 3.94 after 72 h. Nevertheless, the pH for the PP shale revealed a slight increase from pH 5.23 to 5.64 and provided an average value of 5.49 at the end of the experiment.

Figure 3

Variation of solutions pH during the reaction.

The results of pH and mineral composition changes in PR and PP shale suggested that different geochemical reactions occurred between shale and slickwater fracturing fluid. Previous studies confirmed that pyrite oxidation easily occurred in the environment of the atmosphere or water-containing oxygen.[22,40−42] The dissolved oxygen and the oxidative gel breakers in the fracturing fluid could have substantial effects on pyrite dissolution. The reactions of pyrite oxidation are represented as followsBased on the results of FE-SEM, XRD, and solution pH, pyrite oxidation occurred in the PR shale during the experiment. The pyrite oxidation released H+, which resulted in a sharp decrease in the solution pH after the reaction of slickwater and PR shale during the first 24 h, and a gradual increase in the pH toward the end indicated that pH buffering took place. The XRD results of PR shale showed that the calcite was not detected after the reaction, indicating that the acid produced by pyrite oxidation was gradually consumed by the dissolution of calcite. The geochemical reaction of PR shale may be shown as followsOwing to the main mineral of quartz and clay for the PP shale, the only geochemical reaction of the clay swelling occurred during the experiment.

Pore-Structure Analysis

Characteristics of N2 Adsorption–Desorption Isotherms

Low-temperature N2 adsorption–desorption isotherms of the unreacted and reacted PR and PP shale samples are shown in Figure . The hysteresis loops in N2 isotherms were observed when the relative pressure (P/P0) was higher than 0.45, reflecting that the capillary condensation occurred in the macropores (>50 nm) and mesopores (2–50 nm).[43] The hysteresis loops of untreated and reacted PR and PP shale were small, and the isotherms increased to infinity when P/P0 = 1, which further reflected the existence of macropores and mesopores in these shales.[44] Based on the IUPAC recommendation, the isotherm shapes and the hysteresis loops of the unreacted and reacted shale samples belonged to type IV and H2, respectively. The characteristics of N2 adsorption–desorption isotherms indicated that the PR and PP shale possessed a continuous and open nanometer pore system, and pores in these shales were mainly slit pores and ink-bottle pores.[44,45]

Figure 4

N2 adsorption–desorption isotherms for the PR (a) and PP (b) shale sample.

As shown in Figure , some differences in the quantity of adsorbed N2 at the highest pressure appeared for the PR and PP shale. The maximum amounts of adsorbed N2 for the PR shale increased from 3.01 to 9.10 cm3/g after reaction with slickwater fracturing fluid (Figure a). However, a decrease in the maximum quantity of adsorbed N2 from 10.37 to 7.85 cm3/g was observed for the PP shale (Figure b). The phenomenon illustrated that many nanopores were formed in the PR shale because the adsorption capacity of shale mainly depended on micropores (<2 nm) and mesopores (2–50 nm).[46] However, the nanopores were reduced in the PP shale after reaction with the slickwater fracturing fluid.

Changes in the Pore-Structure Parameters

The specific surface area (SSA), pore volume (PV), and average pore diameter (Ra) of unreacted and reacted shale samples were analyzed by the N2 adsorption–desorption measurement, and the results are illustrated in Table . For the PR shale, the total SSA (TSSA) and total PV (TPV) increased greatly by 5.4 times and 3.1 times after reaction with fracturing fluid, respectively, but Ra decreased (Table ). However, the TSSA, TPV, and Ra of sample PP showed completely opposite characteristics. The decreasing rates of TSSA and TPV were 33.2 and 22.9%, respectively, and the Ra increased from 5.976 to 6.893 nm. A previous study demonstrated that when the formation of new small pores was larger than the transformation degree of small pores to large pores, the Ra would show a decreasing trend.[31] To deeply explore the changed characteristics of micro-, meso-, and macropores before and after reaction with fracturing fluid, the percentage of each type of pore was calculated (Figure ). The unreacted PR and PP shale mainly contained meso- and macropores, which provided more than 90% of the TPV. After slickwater treatment, the SSA and PV of the mesopores increased by 8.86 and 5.47% in PR shale, respectively. However, the SSA and PV of the mesopores decreased by 0.93 and 11.73% for the PP shale after reacting with fracturing fluid, respectively. These results illustrated that the variation in mesopores played a major role in the characteristics of pore structure change after slickwater treatment.

Table 3

Pore-Structure Parameters of the Unreacted and Reacted Shale Samplesa

		SSA (m2/g)				PV (cm3/g)
samples	state	micro-	meso-	macro-	TSSA	micro-	meso-	macro-	TPV	Ra (nm)
PR	unreacted	0.378	1.061	0.050	1.489	0.00018	0.00254	0.00194	0.00466	12.521
	reacted	1.373	6.407	0.220	8.000	0.00062	0.00854	0.00507	0.01423	7.115
PP	unreacted	2.245	7.559	0.061	9.865	0.00105	0.01169	0.00200	0.01474	5.976
	reacted	1.438	4.990	0.164	6.592	0.00067	0.00768	0.00301	0.01136	6.893

Note: micro-, meso-, and macro- represent micropores (<2 nm), mesopores (2–50 nm), and macropores (>50 nm), respectively. TSSA and TPV represent total SSA and total PV, respectively.

Figure 5

Percentages of micro-, meso-, macropores of SSA (a) and PV (b) before and after slickwater treatment.

Evolution of Pore Area and Volume Distribution

To better reflect the distributions of SSA and PV for the PR and PP shale, their cumulative and incremental values were investigated. It could be seen that the cumulative pore area (Figure a) and pore volume (Figure c) gradually increased with the increase in the pore diameter of the PR shale. There was an obvious increase in cumulative pore area in the interval of 2–10 nm. Combined with the distribution characteristics of incremental pore area, it was easy to find that the pore area of the reacted PR shale began to increase after 2 nm and had a peak at about 3 nm (Figure b). The results reflected that changed characteristics of pore area in the reacted PR shale were attributed to the increase in pores mainly occupied by mesopores with a diameter of 2–10 nm. Moreover, the pore volume of the PR shale experienced a significant increase after treatment with slickwater fracturing fluid (Figure c), which exhibited a trimodal feature at about 9.2, 50, and 132.9 nm (Figure d). These changes illustrated that the interactions between fracturing fluid and PR shale caused a more serious increase in pore volume especially the meso- and macropores, which was consistent with the analysis of pore-structure parameters. According to the previous research report, the pore area was mainly determined by small pores, while the pore volume was mainly by large pores;[31] thus, the pore area increased with the growing number of mesopores, and the pore volume increased with the growing number of meso- and macropores for the reacted PR shale.

Figure 6

Pore area and volume distributions of the unreacted and reacted PR shale: (a) cumulative pore area; (b) pore area distribution corresponding to average pore diameter; (c) cumulative pore volume; and (d) pore volume distribution corresponding to average pore diameter.

The pore area and volume distributions of the unreacted and reacted PP shale are displayed in Figure . After treatment with the slickwater fracturing fluid, the cumulative pore area (Figure a) and volume (Figure c) of the PP shale experienced a decrease. The incremental pore area of the reacted PP shale was significantly lower than that of unreacted in the range from 2 to 10 nm (Figure b). In addition, the variation in the incremental pore volume was concentrated in the pore interval below 10 nm (Figure d). These results indicated that the slickwater treatment mainly altered the mesopore structures of 2–10 nm, whereas it had little effect on the macropores of the PP shale.

Figure 7

Pore area and volume distributions of the unreacted and reacted PP shale: (a) cumulative pore area; (b) pore area distribution corresponding to average pore diameter; (c) cumulative pore volume; and (d) pore volume distribution corresponding to average pore diameter.

Evolution of Pore Fractal Dimensions

According to eqs and 2, the fractal parameters and fitting curves were obtained, as illustrated in Table and Figure . Owing to the fact that the gas adsorption mechanism at relative pressure below 0.45 was mainly monolayer or multilayer sorption, while at relative pressure above 0.45 was mainly capillary condensation.[47] Therefore, the D1 and D2 were calculated as the P/P0 of 0–0.45 and 0.45–1 (Table ). D1 and D2 showed a positive correlation with the pore surface roughness and pore structural complexity, respectively. The space structural irregularity and heterogeneity of the pore surface and pore internal space could be quantitatively characterized.[48,49]

Table 4

Fractal Dimensions of D1 and D2 of the Unreacted and Reacted Shale Samples

		P/P0: 0–0.45		P/P0: 0.45–1
samples	state	D1	R2	D2	R2
PR	unreacted	2.507	0.962	2.664	0.997
	reacted	2.492	0.995	2.782	0.983
PP	unreacted	2.633	0.989	2.824	0.977
	reacted	2.620	0.988	2.785	0.982

Figure 8

Fractal dimension features of PR (a: unreacted; b: reacted) and PP (c: unreacted; d: reacted) shale samples treated by slickwater fracturing fluid.

As shown in Table , the correlation coefficients (R2) of all of the samples were large (>0.96), suggesting that the fractal dimensions of data fitting were reliable. Furthermore, the values of D1 and D2 ranged from 2 to 3, indicating that the pore structure of unreacted and reacted shale has good fractal properties. It can be seen that the D1 of the PR and PP shale was slightly reduced after slickwater treatment, which reflected that the surface roughness of pores did not show any notable change. However, the D2 variation of the PR and PP shale showed a different feature. For PR shale, the D2 increased from 2.664 to 2.782, indicating that the slickwater treatment was able to enhance the pore structural complexity. For the PP shale, the D2 decreased from 2.824 to 2.785, illustrating that the pore structure became more homogeneous after the reaction. Previous studies demonstrated that D2 was positively related to the total specific surface and pore volume, while negatively related to average pore diameter.[31,50] As discussed above, the TSSA and TPV of the PR shale increased and Ra decreased after slickwater treatment, which was induced by the increase in mesopores. Thus, the growth in the number of mesopores in the PR shale led to an increase in the pore-structure complexity and a weakening in the pore connectivity. For the PP shale, the variations in D1 and D2 displayed good accordance with the pore-structure parameters, and the reduction of TSSA and TPV caused by clay swelling led to the decrease of pore surface roughness and pore-structure complexity.

Mechanism of Pore-Structure Change during Hydraulic Fracturing

According to the analysis results of the FE-SEM and XRD, the slickwater fracturing fluid can cause the oxidation of pyrite and the swelling of clay minerals. Therefore, the changes in pore structure were related to multiple chemical reactions. A schematic diagram of the reactions and results during hydraulic fracturing is illustrated in Figure . Through comparison of the pyrite-rich and pyrite-poor shales, we found that the pyrite oxidation occurred, and then formed many dissolution pores and released H+. The unstable minerals such as calcite in the shale could dissolve in the acidic solution, transforming micropores to meso- or macropores. Jew found that the release of Fe showed a steady increase in the carbonate-poor shale, while little to no Fe was detected in the carbonate-rich shale during 3 weeks of reaction,[5] which illustrated that the content of carbonates in shale had an obvious impact on pyrite oxidation. Due to the lower carbonate minerals of PR shale used in this study, pyrite dissolution occurred. Moreover, we found that the pyrite oxidation can cause an extensive increase in the mesopores in pyrite-rich shale, and the release of H+ would also cause the dissolution of carbonate minerals, which thereby increased the TSSA and TPV and enhanced the pore-structure complexity. Therefore, it can be inferred that the pyrite oxidation during hydraulic fracturing can improve the number of mesopores in the shale gas reservoir, thereby having a favorable impact on gas production.

Figure 9

Schematic diagram of the reactions and results during hydraulic fracturing.

Meanwhile, the clay swelling could not be neglected during hydraulic fracturing. Previous studies illustrated that the illite/smectite, illite, and chlorite were the dominant clay minerals in the shale of Niutitang Formation and Xiamaling Formation.[29,51−53] The PR and PP shale containing clay minerals, which were dominated by swelling illite–smectite mixed-layer minerals, could easily lead to water-sensitive damage and hydration expansion after the fracturing operation.[38] The clay swelling could block the pore throats and reduce the connection between nanopores, which were identified by the results of the pore structure change of the PP shale. Therefore, pyrite oxidation and calcite dissolution caused by fracturing fluid treatment ultimately contributed to the increase in the number of nanopores especially the mesopores for PR shale, and the clay swelling led to the reduction in the number of nanopores for the PP shale.

Conclusions

In this study, by means of FE-SEM, XRD, and low-temperature N2 adsorption, the mineral composition and nanopore structure characteristics evolution of two shales with different pyrite contents were examined before and after treatment of slickwater fracturing fluid. The major conclusions of this work are as follows:

The FE-SEM and XRD results illustrated that the pyrite dissolved significantly in the PR shale after slickwater treatment, and the pyrite oxidation caused the pH of the solution to decrease, which further led to calcite dissolution. However, the main chemical reaction of the PP shale was clay swelling. The different reactions of shale minerals then caused a change in the pore structure.

According to the low-temperature N2 adsorption measurement, the TSSA and TPV significantly increased, while Ra decreased for the PR shale after the reaction. Nevertheless, the TSSA, TPV, and Ra of the PP shale showed completely different characteristics. The dissolution of pyrite and calcite resulted in an increase in the number of mesopores and the pore-structure complexity of the PR shale. The clay swelling reduced the pore volume and pore area of the PP shale and then led to the decrease in pore surface roughness and pore-structure complexity.

Compared with the result of pore-structure characteristic of shale with different pyrite contents under fracturing operation, the mechanism of the pore structure change was revealed. The change in pore structure was related to the pyrite oxidation, calcite dissolution, and clay swelling. Moreover, this study documents sound evidence that pyrite oxidation has a favorable impact on the pore structure of shale during hydraulic fracturing.