A preliminary study of oxidant stimulation for enhancing coal seam permeability: Effects of sodium hypochlorite oxidation on subbituminous and bituminous Australian coals

International Journal of Coal Geology 200 (2018) 36–44

Contents lists available at ScienceDirect

International Journal of Coal Geology journal homepage: www.elsevier.com/locate/coal

A preliminary study of oxidant stimulation for enhancing coal seam permeability: Effects of sodium hypochlorite oxidation on subbituminous and bituminous Australian coals

T

Zhenhua Jinga, Shilo A. Mahoneya, Sandra Rodriguesb, Reydick D. Balucana, Jim Underschultzc, ⁎ Joan S. Esterleb, Thomas E. Rufforda, Karen M. Steela, a b c

School of Chemical Engineering, The University of Queensland, St Lucia 4072, Australia School of Earth and Environmental Sciences, The University of Queensland, St Lucia 4072, Australia Centre for Coal Seam Gas, The University of Queensland, St Lucia 4072, Australia

A R T I C LE I N FO

A B S T R A C T

Keywords: Coal seam gas Coal porosity Oxidant stimulation Coal structure Oxidation mechanism

Chemical oxidation is proposed as an effective means to react and dissolve small regions of coal in the near wellbore region, thereby raising permeability for gas flow. In this study, we investigated the effect of sodium hypochlorite (NaClO) treatment on the structure of bituminous coal (Coal B) and subbituminous coal (Coal S) separately from the Bowen and Surat basins in Queensland, Australia. Swelling and leaching tests showed that both coals swelled, dissolved and broke in 5%wt. aqueous solutions of NaClO. Coal S reacted more vigorously in 5% NaClO, with 49% mass loss and 3840 mg/L of dissolved organic carbon (DOC) measured in the oxidant filtrate, than Coal B. The Coal B mass loss in 5% NaClO was 4.5% with 430 mg/L DOC measured in the filtrate. After NaClO treatment the total accessible pore volume of Coal S particles increased from 4.6% to 6.1%, and the porosity of Coal B increased from 8.6% to 8.9%. Pore size distributions determined from mercury intrusion porosimetry (MIP) indicated that oxidation enlarged the pores in Coal S more significantly than Coal B. Scanning electron microscopy (SEM) confirmed oxygen generated large pores on the surface of Coal S particles, but there were no significant changes on Coal B. We used a microfluidic cleat flow cell (CFC) to inject NaClO into artificial channels scribed in polished samples of Coal S and Coal B, and measured an increase in the widths of the channels after NaClO treatment. The increase in channel width observed in the CFC indicated that coal solubilisation was a more dominant mechanism than coal swelling. Similarly, the channel aperture of Coal S increased more than Coal B. CFC results also showed that NaClO etched dull coal bands (inertinite-rich) more significantly than bright coal bands (vitrinite-rich), and we proposed this result was due to the greater porosity in semi-fusinite, which allowed greater penetration of NaClO in dull coal bands than in bright coal bands. The low coal rank sample (Coal S) with higher liptinite content and more oxygen content was more susceptible to oxidisation by NaClO than Coal B.

1. Introduction Many low permeability coal seam gas (CSG) reservoirs require stimulation to allow economic development of the play, even if the seams contain high gas contents (Gamson et al., 1996; Laubach et al., 1998; Liu and Harpalani, 2013; Moore, 2012). Various CSG stimulation techniques have been used commercially, with the most common techniques including hydraulic fracturing, cavity well completions, and horizontal wells (Mavor and Robinson, 1993; Palmer, 1992; Palmer, 2010). However, there are problems associated with these methods, such as damage to the cleat networks through shear failure, creation

⁎

Corresponding author. E-mail address: [email protected] (K.M. Steel).

https://doi.org/10.1016/j.coal.2018.10.006 Received 19 February 2018; Accepted 17 October 2018 Available online 17 October 2018 0166-5162/ © 2018 Published by Elsevier B.V.

and transport of fines, and in the case of hydraulic fracturing the need for large volumes of water to be delivered at high pressures (Chen et al., 2006; Magill et al., 2010; Mavor and Robinson, 1993; Olsen et al., 2007; Palmer et al., 2005). Consequently, there is a need for research and development of new effective and efficient stimulation methods to stimulate low permeable coals. Acid stimulation to increase coal permeability by demineralization in cleats has been reported (Balucan et al., 2016; Turner and Steel, 2016; Turner et al., 2013). Alternatively, we have recently reported oxidant stimulation of coal seams (Jing et al., 2018), and in this paper, we further evaluate the potential of NaClO to increase cleat porosity and enhance coal permeability by dissolving coal

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

matrix. Jing et al. (2018) used unconfined coal particles (3–4 mm in diameter) to study the coal behaviours when oxidants come into contact with coal. Based on the coal behaviours, different oxidant candidates were screened including Sodium Hypochlorite (NaClO), Potassium Permanganate (KMnO4), Hydrogen Peroxide (H2O2) and Potassium Persulfate (K2S2O8). NaClO was determined as the most effective oxidant given its high reactivity with coal (Jing et al., 2018). In NaClO, unconfined coal particles presented different behaviours including solubilisation, swelling, and breakage (Jing et al., 2018). However, the net effect of these behaviours on coal cleats and coal permeability was unclear (Jing et al., 2018). Coal solubilisation by NaClO might enlarge coal cleat aperture, while swelling could narrow it. The effect of breakage caused by dissolution and swelling was indiscernible, as breakage might generate new fractures in coal seams increasing fracture system connectivity. On the other hand, coal breakage might lead to coal fines generation that can block the fractures thus decreasing the permeability. Therefore, this paper reports the direct effects of oxidants on artificial coal cleats within a cleat flow cell (CFC) that enables cleat aperture to be measured before and after treatment. While the coal cleats play a significant role in contributing to the permeability (Flores, 2014; Gamson et al., 1996; Laubach et al., 1998; Levine, 1996; Mazumder et al., 2006; Van Krevelen, 1993), matrix permeability may also be important (Moore, 2012). How these two types of permeability interact with each other determines the overall gas production characteristics for a well (Flores, 2014; Laubach et al., 1998; Moore, 2012). The matrix permeability is determined primarily by the connectivity and distribution of the pore system within the coal matrix, especially macro- and mesopores (Moore, 2012). It has been shown that higher diffusivity was generally related to greater macropore porosity in coals (Gamson et al., 1996). Gamson et al. (1996) also suggested that the presence of open and continuous microstructures (micro-fractures in bright coal and micro-cavities in dull coal) might enable laminar flow to begin before gas enters the cleat. Therefore, if the process of coal oxidation increases the pore system connectivity or generates “micro-structure” in the matrix by dissolution, it might be of particular importance with respect to total permeability and overall gas production. Additionally, this stimulation method only targets small areas in the near wellbore region, so it is not likely to influence future mining activities. The net effectiveness of oxidation on coal seam gas stimulation may be contingent on either the coal rank or maceral composition or both. Low rank coals have previously been proposed to be easily oxidized by NaClO. For example, Liu et al. (2013a) reported that low rank lignite (C% = 70.84 wt%, d.a.f.) was easily oxidized in a 6% NaClO solution with > 85.3 wt% water soluble organic matter obtained after oxidation. In contrast, Wang et al. (2014) obtained with the same method only 42 wt% water soluble organic substances after oxidation of anthracite (C% = 88.40 wt%, d.a.f). Furthermore, the oxidative products in each experiment were different with mostly alkanoic acid and alkanedioic acids produced in lignite oxidation and benzene carboxylic acids from anthracite oxidation (Liu et al., 2013a; Liu et al., 2013b; Wang et al., 2014). The difference in oxidative products from different rank coals is due to different coal structures with lignite being more aliphatic and having more oxygen-containing structures than anthracite (Chen et al., 2012; Ibarra et al., 1996; Liu et al., 2016; Lv et al., 2014; Murata et al., 2001). Oxygen-containing compounds abundant in low rank coals (Arenillas et al., 1999) were reported to react easily with NaClO (Yu et al., 2014). In terms of coal maceral composition, liptinite generally exhibits the lowest aromaticity and the longest aliphatic chains, while inertinite has the highest aromaticity and polycondensed structures, and vitrinite generally exhibits intermediate characteristics between them (Chen et al., 2012; Van Niekerk et al., 2008). Furthermore, liptinite contains a broader range of oxygen-containing groups than other macerals (Guo and Bustin, 1998). Thus, liptinite was expected to be favourable for NaClO oxidation according to the NaClO reactivity with

Table 1 Ultimate and proximate analysis of coal drill core samples used in this study. Coal Sample

Coal B Coal S

Ultimate analysis

Proximate analysis

(wt%, d.a.f)

(wt%, d.b)

(wt %, ar)

C

H

S

N

O (diff.)

Ash

VM

FC (diff)

MC

85.54 77.88

5.06 6.15

0.33 0.57

1.97 1.31

7.10 14.08

6.6 12.3

28.7 43.1

62.2 39.5

2.5 5.1

C – carbon; H – Hydrogen; S – Sulphur; N – Nitrogen; O – Oxygen; VM – Volatile Matter, FC – Fixed carbon; MC – Moisture content; d.a.f: dry ash free base; d.b: dry base; ar: as received.

coals described in previous research (Chakrabartty and Kretschmer, 1972; Liu et al., 2016; Mayo, 1975; Yao et al., 2010). Therefore, in this preliminary study, two coals, with different rank and maceral composition, from CSG production basins (Bowen and Surat basins) in Queensland, Australia, were used to investigate the effects of NaClO oxidation on coal pore and cleat structures. The relationship between these structure changes and the coal permeability is discussed. 2. Methodology 2.1. Coal samples and characterisation High volatile A bituminous (Coal B) and subbituminous (Coal S) coal drill core samples were collected from the Bandanna Formation in Bowen Basin and the Walloon Subgroup from Surat Basin, respectively (Fielding et al., 1995). The chemical and petrographic analyses of each sample are summarised in Table 1 and Table 2. The ultimate and proximate analyses of the coal samples were carried out at the ALS Coal Division, in Richlands, Queensland, Australia. Elemental oxygen contents were determined by difference (Table 1). Random vitrinite reflectance (RO) and the maceral composition were determined at the School of Earth and Environmental Sciences laboratory, The University of Queensland (Table 2). Sample blocks were ground, polished and examined under a Leica DM6000 microscope® fitted with an internal 10× lens and a 50× reflected light oil immersion objective, giving a total of 500× magnification. For imaging a high resolution black and white camera and a colour camera are attached to the microscope. The former is also used for reflectance analysis according to ISO 7404-5 (2009). Hard- and software for this microscope system were made by Hilgers Technisches Buero (Diskus Fossil). The maceral groups were quantified by a 500 point-count technique following the procedures described in the standard ISO 7404-3 (2009). 2.2. Swelling test and leaching test A time-lapse photographic method combined with image analysis was conducted to study the different coal swell/shrink behaviours in 5%wt. NaClO, while leaching tests were used to investigate the mass Table 2 Petrographic analysis based on maceral groups of coal drill core samples used in this study. Coal sample

Coal B Coal S

37

Petrographic analysis (vol%)

Vitrinite reflectance

Vitrinite

Inertinite

Liptinite

Mineral Matter

Ro, %

62.8 56.4

32.4 3.6

2.2 29.0

2.6 11.0

0.84 0.47

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

change and dissolved organic carbon (DOC) of the filtrate recovered after coal oxidation. In the swelling test, individual dried coal particles (3–4 mm in diameter) were placed in 5 mL quartz glass tubes and mounted against a well-lit white background at room temperature (22 ± 1 °C) and under ambient pressure (1 atm). The experiment involves seven particles in seven tubes with 4 mL of solution, including one blank test with 4% KCl as the solution. Digital images were captured at 5 min intervals over 30 days for both coals. Image analysis was conducted using Image-J Version 1.46r software which saves the images as a video to examine the reaction phenomenon such as colour change, gas generation and particle movement. Where particle movement occurred, the particle was removed from the analysis. As each individual image is associated with time, a swelling ratio (SR) can be determined by dividing the projection area of the treated coal particle at any particular time step by the projection area at the start. A value of SR = 1 corresponds to no particle size change; SR < 1 indicates particle shrinkage; and SR > 1 indicates particle swelling. Details about the experimental procedures were described previously by Jing et al. (2018). The swelling test is a valuable method not only to examine coal particle size change but also to visualise the coal particle behaviours in oxidants. For instance, the solution colour variation in the swelling test could give clues about the reaction occurring during the experiment, and the breakage of coal particles could illustrate physical effects of oxidation on coal structures (Jing et al., 2018). Leaching tests involved reacting 1 g of dried coal powders (75–212 μm) in 100 mL of solutions (blank test/oxidants) for 24 h at room temperature (22 ± 1 °C) and under ambient pressure (1 atm) with continuous stirring at 300 rpm. At the conclusion of the tests, the coal particles and leachate solution were separated via vacuum filtration using filter paper with a pore size of 5 μm (Jing et al., 2018).

change in the width of artificial channels in coals before and after oxidation by flowing oxidant through scribed channels. Coal cubes (each 15 mm length) were cut perpendicular to the bedding plane and mounted in epoxy blocks. The mounted samples were ground and polished for incident light microscopy according to the procedures in Australian Standard AS2061 (1989). Following the polishing, coal samples were scribed with a 100 μm tungsten-carbide cutting tool at an angle of approximately 55° to ensure that only the sharp corner edge of the cutting tool would scribe the coal surface. The cutting tool was moved by hand control across the coal block at speeds of approximately 5–10 mm/s. Cutting speed within this range has been found to produce channels with consistent depth and width (Mahoney et al., 2017). For the scribed channel, an optical image was first collected with a Leica DM6000 light microscope® equipped with a Leica DFC365 FX Digital high-speed camera®. Image analysis was performed with the Leica Application Suite Advanced Fluorescence software package®. Channel apertures at points of interest were measured with image analysis software Image-J Version 1.46r®. Due to the fine polish of coal, the maceral cut by the artificial channel could be identified on the coal sample surface from the microscope image. After the initial image was obtained, the top of the coal channel was sealed with a polyolefin film. The oxidant was injected into the channels using a 10 mL syringe pump connected with Tygon tubes to inlet and outlet holes (1.5 mm in diameter), as described in Mahoney et al. (2015) and Mahoney et al. (2017). The 5% NaClO solution was injected at a rate of 10 μL/min to reduce the physical effect of fluid flooding on the channel structure. After injection, the sealing film was peeled off and channel images after oxidation were taken with the camera. Based on the images before and after oxidation, the channel apertures and shapes at multiple points along the channel were compared. A set of channels was also flooded with 4% KCl to provide a reference experiment without oxidation.

2.3. Porosity characterisation

3. Results

The bulk or apparent density (ρHg) of the coal samples was measured by mercury intrusion porosimetry (MIP, Micromeritics PoreSizer 9320) at pressure ranging from 1 to 6000 psi and the skeletal density (ρHe) by helium pycnometry (Micromeritics AccuPyc II 1340). The total accessible coal porosity (∅) was calculated using Eq. (1):

∅ (%) = (ρHe − ρHg )/ ρHe × 100

3.1. Coal properties The ultimate analysis shows that Coal B contains a higher carbon content (85.54%) but relatively lower oxygen content (7.10%) than Coal S (correspondingly 77.88% and 14.08%). The petrographic analysis showed Coal B has higher rank (Ro, % = 0.84) than Coal S (Ro, % = 0.47). In terms of the maceral compositions, the two coals have comparable vitrinite content (62.8% for Coal B and 56.4% for Coal S), but Coal B contains significantly more inertinite (32.4% and 3.6% for Coal B and Coal S, respectively) while Coal S is significantly more liptinite rich (29.0% for Coal S and 2.2% for Coal B).

(1)

As coal porosity change could not be measured on the same coal particles, because MIP is a destructive test, a total of 10 g (3–4 mm coal particles) from each coal core was divided into four groups by coning and quartering. Of these, two were then selected for treatment with 5% NaClO and 4% KCl for 7 days. After treatment, the samples were filtered and dried in an oven at 110 °C for 12 h. The porosity tests were repeated twice to assess reliability.

3.2. Swelling test

2.4. Scanning electron microscopy (SEM) imaging before and after oxidation

For both Coal B and Coal S, swelling tests were conducted as shown in Fig. 1, where the green line shows Coal B swelling ratio (SR) and the red line shows Coal S SR. For each coal, the swell test procedure was divided into 3 stages based on the degree of limpidity of the solution. In stage I, the solution was limpid where the coal outline could be accurately distinguished by visual examination. For Coal B, this stage lasted an average of 9 days and coal swelling (SR = 1.5) could be observed. However, for Coal S, this stage was less than one day with an indiscernible SR occurring due to the solution around the particle becoming dark. The dark colour provides evidence of the reaction between the coal and the oxidant, producing molecules from the organic coal structure that are soluble and diffuse into the solution (Fig. 1). Stage II is marked as a brown colour stage where the solution becomes dark. In this stage, the optical recognition of the particle boundary is diminished, thus the SR was less accurate, as represented by dash line for each coal. For Coal S, the sudden decrease in SR after approximately 2 days is

Two groups of coal particles (3–4 mm) were picked from each of the coal cores and were saturated in 5% NaClO and 4% KCl for seven days, respectively. After treatment, the samples were rinsed with distilled water and dried at 110 °C for 12 h. The dried coal particles were mounted on aluminium stubs and sputter-coated with a layer of iridium using a Q150 TS sputter coater® (Quorum Technologies, UK). The coal particles were then viewed using SEM JSM-7001F® at 15 kV in secondary electron mode. 2.5. Cleat Flow Cell (CFC) To investigate the oxidant effect on coal cleats, we used the cleat flow cell (CFC) reported by Mahoney et al. (2017) to observe the 38

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

B and Coal S as well as the blank test based on both coals. The blank tests were used to investigate artifacts of stirring and filtration on experiment results. Artifacts were proven to be negligible with only a slight mass change and negligible DOC in the blank test compared with those in 5% NaClO. Coal B exhibited 4.5% mass loss and 430 mg/L DOC and Coal S showed 49.4% mass loss and 3840 mg/L DOC in the filtrate. Note that the solubilisation of the coal was based on powdered samples, with a large particle size surface area. 3.4. Porosity characterisation of the coals Coal porosity changes were observed for both coals as shown in Fig. 3a. The porosity of Coal B increased averagely from initial 8.6% to 8.9%, while the porosity of Coal S rose from 4.6% to 6.1%. In addition to the increase in total porosity, the pore size distribution (PSD) derived from MIP showed the enlargement of measured pores after oxidation. Fig. 3b and Fig. 3c show the pore size distribution for Coal B and Coal S, respectively, both before (dashed lines) and after (solid lines) oxidation. For Coal B, the number of pores in the size ranging from 0.008 μm (8 nm) to 0.05 μm (50 nm) declined after oxidation, while the number of pores in the size range between 0.05 μm to 5 μm increased, suggesting that some smaller pores (0.008 to 0.05 μm) were enlarged due to oxidation. For pores larger than 5 μm, there was a slight increase in size. Comparatively, there was an increase in the number of pores for Coal S across a larger pore size range (from 0.01 to 100 μm), with no decline of any specific pore size range after oxidation.

Fig. 1. Swell/shrink behaviours of different coal sample particles in 5% NaClO. SR: swelling ratio; IB, IIB, IIIB: Coal B behaviour stages, Is, IIs, IIIs: Coal S behaviour stages.

due to the solution being replaced (through careful siphoning) by fresh oxidant solution. Solution replacement was repeated on two further occasions. Solution replacement was not necessary for Coal B as the solution naturally returned to a limpid state. Therefore, during stage III, the solution becomes clearer again due to the remaining NaClO decomposing the initial oxidation products and reaction approaching equilibrium. The solution colour changing from dark to light during oxidation was also observed previously (Liu et al., 2013a; Miura et al., 1996). In stage III, although the solution was not perfectly clear, it is clear enough to observe an obvious outline of the particles. The final SR for Coal B and Coal S were 2.0 and 0.73, respectively. It could be concluded from the images that swelling, breakage and dissolution are occurring where dissolution and then breakage appears to be dominating for Coal S (SR < 1), while swelling and breakage appears to be dominating for Coal B (SR > 1). It should be noted that Coal S was exposed to more oxidant due to the solution refreshing. Therefore, a direct comparison of the behaviour for the two coals cannot be made.

3.5. Scanning Electron Microscopy (SEM) Comparing Fig. 4a and Fig. 4b by SEM, Coal B surface was generally roughened by oxidation, changing from a smooth morphology to a mottled surface but no obvious new pores or fractures with magnification of 1000. However, for Coal S, a more dramatic change occurred to the surface topography (Fig. 4c and d). Coal pores explicitly appeared after oxidation from what was previously a smooth surface. On a smaller scale, skeletal parallel structures in the top right to bottom left direction could be detected. 3.6. Cleat Flow Cell (CFC) The artificial cleat aperture and the change of cleat shape were examined before and after coal oxidation for both coals. Firstly, to investigate the physical influence of fluid injection on channel aperture and shape, blank tests with 4% KCl were conducted prior to oxidation as shown in Fig. 5 and Fig. 6 for Coal B and Coal S, respectively. The photomicrographs show that the changes in both channel aperture and shape were indiscernible, regardless of maceral composition along the channel. After the blank test, 5% NaClO was used to flow through the channels for both Coal B and Coal S. Fig. 7a and Fig. 7b show the images zoomed in to a section of the channel of Coal B before and after 5% NaClO oxidation. The results illustrate that oxidation effect on channels varied between coal macerals. The channel aperture in the inertinite bands increased 56.8% from 322 μm to 505 μm, while the aperture in the vitrinite bands only increased 3.6% from average 631 μm to 658 μm. The channel shape also changed significantly in the inertinite bands but indiscernibly in vitrinite bands. Comparing the Coal S channel post-oxidation (Fig. 8b) with preoxidation (Fig. 8a), the channel aperture increased 118% from an average width of 350 μm to 762 μm, which was much higher than that for Coal B. Moreover, all the initial sharp angles on the Coal S channel wall were altered to result in reducing the tortuosity of the channel (Fig. 8b). Note that two new channels were also generated roughly perpendicular to the original artificial channel (Fig. 8b). The new channel on the right occurs in an inertinite-rich band (semi-fusinite). Semi-fusinite

3.3. Leaching test The colour change in the swelling test qualitatively indicates coal solubilisation in 5% NaClO, while the leaching test demonstrates the reaction quantitatively by measuring the coal sample mass loss as well as the DOC in the filtrate. Fig. 2 shows the leaching test results for Coal

Fig. 2. Mass loss and DOC in the filtrate of different coal samples after oxidation by 5% NaClO. 39

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

Fig. 3. Total porosity increase (a) and pore size distribution (PSD) variation of Coal B and Coal S (b and c, respectively) after oxidation by 5% NaClO. The PSD tests were conducted twice for each coal sample named I and II.

4. Discussion

is known to be porous (Flores, 2014; Harris and Yust, 1976) and perhaps the oxidant penetrated that porosity leading to enhanced reactive dissolution. The new channel on the left followed an original fracture located in the inertinite rich band (Fig. 8a), suggesting the combination effects that the oxidant penetrated the original fracture and reacted with inertinite.

In this study, we found that swelling of coal particles was the dominant mechanism for changes observed in Coal B in 5% NaClO, and that solubilisation was the dominant observed in Coal S. The Coal B solute reached a dark brown colour more slowly than Coal S. The swelling test observations could be complementarily explained by the leaching test. Coal B measured mass loss and DOC of the solute filtrate was 4.5% and 430 mg/L, respectively, and for Coal S, it was 49% and

Fig. 4. SEM images of Coal B and Coal S showing the surface morphology change after oxidation. a. Coal B pre-oxidation; b. Coal B post-oxidation; c. Coal S preoxidation; d. Coal S post-oxidation. Images were taken in secondary electron mode. 40

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

Fig. 5. Photomicrographs of cleat flow cell blank test of Coal B (Ro % = 0.84). a. initial channel; b. channel after 4% KCl injection.

From laboratory experimentation and imaging before and after oxidation, it was observed that total porosity of both coals increased after oxidation, especially for Coal S with total porosity increasing by 32.6%. The total porosity increase was demonstrated by the PSD change, where the coal pores were enlarged by oxidation. SEM results visualised this enlargement in pores and the generation of new pores, especially on Coal S surfaces (Fig. 4d). In terms of the cleat aperture, CFC tests showed the etching of cleats resulting in increased aperture for both coals. Specifically, the average channel aperture of Coal B increased by 56.8% in inertinite bands and 3.6% in vitrinite bands, while it increased by > 118.0% in Coal S, particularly in the areas with new fractures developing perpendicular to initial channels. The generation of new channels in Coal S is of particular interest. Firstly, the right channel (Fig. 8b) occurs in an inertinite-rich (semi-fusinite) band. Secondly, the left channel (Fig. 8b) followed an original fracture located in an intertinite-rich band. This illustrates that when the oxidant is injected into the coal seams, it might follow existing fractures and increase the fracture aperture. Therefore, both the porosity and CFC tests showed the potential of coal oxidation to enlarge or even generate new microstructures as

3840 mg/L, respectively. The swelling and leaching test results suggest that Coal S is more reactive to NaClO oxidation than Coal B. The high solubilisation (49.4% in coal powder samples) and low swelling ratio (< 1) for Coal S in 5% NaClO illustrate that Coal S permeability might be increased after 5% NaClO oxidation. Coal permeability is predominantly controlled by the coal cleat system (Clarkson and Bustin, 1996; Flores, 2014; Laubach et al., 1998; Moore, 2012). Additionally, Gamson et al. (1996) found that between the gas diffusion in micropores and the laminar flow in the cleat system, there is an intermediate level for methane flow through “microstructures” in the coal matrix. They defined these microstructures as micro-fractures which occurred in bright coal and micro-cavities in dull coal. The existence of microstructure might advance laminar flow of gas and significantly influence gas flow in coal seams (Gamson et al., 1996). Therefore, oxidation that induces changes to coal structure may benefit both the bulk permeability by increasing cleat aperture, but also enhance the microstructural architecture of the matrix that delivers desorbed gas to the cleat, including the enlargement of coal pores, an increase in coal cleat aperture, and the generation of new microstructures, as observed in the results.

Fig. 6. Photomicrographs of cleat flow cell blank test of Coal S (Ro % = 0.47). a. initial channel; b. channel after 4% KCl injection. 41

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

Fig. 7. Photomicrographs of cleat flow cell test for Coal B (Ro % = 0.84). a. initial channel; b. channel after 5% NaClO injection.

of being inherited from the original plant structure. In the early stages of coalification, the cell walls (composed by lignin-cellulose) were likely gelified within themselves without destruction of the cell tissue and humic gels precipitated in the open cell lumens. During oxidation with NaClO the humic gels may react with the oxidant agent, but the cell wall being more resistant may remain preserved (Fig. 4d). Furthermore, within the Walloon Subgroup coals (Coal S) the suberinite maceral is usually the main maceral of the liptinite group (Khorasani, 1987). Suberin is a polymer containing polyesters and aromatics (Pickel et al., 2017) and preferentially reacts with the oxidising agent. The NaClO oxidation preference is complex and could be affected by sample molecular structures and/or coal properties which control the accessibility of NaClO to react with coal. In terms of the coal molecular structures, NaClO was reported to preferentially oxidise the long aliphatic side chains and some polynuclear aromatics where oxygen functional groups exist, although uncertainty still exists (Chakrabartty and Kretschmer, 1972; Liu et al., 2016; Mayo, 1975; Yao et al., 2010; Yu et al., 2014). Coal molecular structures vary with coal rank and maceral compositions. Previous studies have found an increase in aromaticity and a loss of aliphatic and oxygen-containing structures with increasing rank (Ibarra and Juan, 1985; Ibarra et al., 1996; Wang and Griffiths, 1985). Additionally, a higher content of organic oxygen in lower rank coals was also reported (Liu et al., 2016; Okolo et al., 2015). Among the maceral groups, inertinite shows the highest aromaticity,

defined by Gamson et al. (1996) in coal to facilitate the gas diffusion in the coal matrix and enhance laminar flow in the coal cleat system. Coal S appears to be more reactive than Coal B. To quantify the permeability enhancement, coal core flooding tests under in situ confining and fluid pressures are suggested for future research. The change in the coal structure after oxidation appears to be due to coal swelling and dissolution. According to the coal associated and nonassociated combined structure model (Van Niekerk et al., 2010), coal consists of a macromolecular structure and a mobile phase (trapped low molecular weight compounds). When the coal macromolecular structure is swollen, those trapped substances in the coal structure may be released which could provide void space in the coal structures, thus increasing the coal porosity. Additionally, NaClO could oxidise coals to generate a series of water-soluble products including benzene carboxylic acids, chloro-substituted alkanoic acids and alkanedioic acids (Liu et al., 2013a; Liu et al., 2013b; Mayo, 1975; Wang et al., 2014). Coal dissolution in NaClO was reported previously (Chakrabartty and Kretschmer, 1972; Liu et al., 2016; Mayo, 1975) and has been demonstrated in this research by the mass loss of coal samples and DOC in the filtrate. The partial dissolution of coal could also lead to porosity increase. The parallel structures of Coal S observed by SEM is direct evidence of the partial oxidation, indicating that the oxidant preferentially reacts with certain parts of the coal. The resulting post reaction texture and morphology highlights coal heterogeneity indicative

Fig. 8. Photomicrographs of cleat flow cell test of Coal S (Ro % = 0.47). a initial channel; b. channel after 5% NaClO injection. 42

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

6. Inertinite was found to be more reactive than vitrinite, which may be due to its porous structure allowing oxidant penetration.

more polycondensation and is more structurally ordered, while liptinite shows lower aromaticity and longer aliphatic chains (Chen et al., 2012; Mastalerz and Bustin, 1993a; Mastalerz and Bustin, 1993b; Van Niekerk et al., 2008). According to the ultimate and petrographic analysis, Coal B is of higher rank with abundant inertinite group macerals, whilst Coal S has more liptinite and oxygen content (Table 1 and Table 2). Therefore, Coal S was characterised as lower aromaticity with abundant long alkyl side chains and methylene bridges. All these characteristics demonstrate the favourability of Coal S for NaClO oxidation. Taking NaClO oxidation preference and the molecular characterisation of different maceral groups into consideration, it should be expected that the inertinite-rich bands in Coal B would be less reactive than vitrinite-rich bands when exposed to NaClO. However, the opposite was observed (Fig. 7b). Actually, the molecular structural similarity of inertinite and vitrinite groups in the same coal rank was reported by Van Niekerk et al. (2008). They found that although inertinite-rich coal was structurally more ordered and more aromatic, the general structures of both coals had similar average aromatic cluster sizes (16 carbons for vitrinite-rich and 18 carbons for inertinite-rich coals) and similar number of cluster attachments (6 attachments for vitrinite-rich and 5 attachments for inertinite-rich coals). Since the chemical-structures cannot explain the oxidative heterogeneity between inertinite and vitrinite bands in Coal B, the different pore systems in these two maceral groups might primarily affect the behaviours. Inertinite was reported as the most porous maceral group (Flores, 2014; Harris and Yust, 1976) and primarily composed of mesopores and macropores associated with the original plant fragments, while vitrinite-rich coal mainly contains micropores (Gamson et al., 1996; Harris and Yust, 1976; Mahoney et al., 2017). The differences in surface morphology between the two maceral groups were also observed by SEM, where the inertinite-rich bands had a rougher texture compared to the smooth surface displayed on vitrinite-rich bands (Gamson et al., 1996; Gan et al., 1972; Mahoney et al., 2017). The larger pores in inertinite bands are favourable for oxidant molecules to penetrate and react with coal surface. In the case of cleat-rich bright coal bands, the cleat system can also favour the oxidation process. Further research will include a study between NaClO and coal samples with different maceral compositions in the same coal rank, as well as coals containing similar maceral compositions but with different ranks. This is of particular importance, because it will help to decide what oxidative method to target in future experiments or field trial.

Based on the results and discussion, Coal S permeability is expected to increase more after NaClO stimulation at in-situ conditions. To examine this further, core flooding tests with both coals and NaClO are recommended to measure time lapse permeability changes with the core under confining and pore pressure conditions matching typical CSG reservoir conditions. The specific effects of coal rank and maceral compositions on NaClO oxidation should be further studied on coal samples with similar maceral composition but different ranks and vice versa. Acknowledgement We gratefully acknowledge the aid of Mr. Jinxuan Zhang, School of Chemical Engineering, for the SEM experiments, Miss Nikola Van de Wetering, School of Earth and Environmental Sciences, for the coal sample preparation and Mr. Michael Tobe, School of Earth and Environmental Sciences, for the DOC analysis. The facilities and technical assistance of the Australian Microscopy & Microanalysis Research facility at the Centre for Microscopy & Microanalysis at The University of Queensland are acknowledged. This work was supported by the Centre for Coal Seam Gas, The University of Queensland and its industry partners APLNG, Arrow Energy, QGC and Santos as well as the China Scholarship Council. References Arenillas, A., Rubiera, F., Pis, J., 1999. Simultaneous thermogravimetric–mass spectrometric study on the pyrolysis behaviour of different rank coals. J. Anal. Appl. Pyrolysis 50, 31–46. Australian Standard AS2061, 1989. Preparation of Coal Samples for Incident Light Microscopy. Standards Australia International Ltd. Balucan, R.D., Turner, L.G., Steel, K.M., 2016. Acid-induced mineral alteration and its influence on the permeability and compressibility of coal. Journal of Natural Gas Science and Engineering 33, 973–987. Chakrabartty, S., Kretschmer, H., 1972. Studies on the structure of coals: part 1. The nature of aliphatic groups. Fuel 51, 160–163. Chen, Z., Khaja, N., Valencia, K., Rahman, S.S., 2006. Formation Damage Induced by Fracture Fluids in Coalbed Methane Reservoirs, SPE Asia Pacific Oil & Gas Conference and Exhibition. Society of Petroleum Engineers. Chen, Y., Mastalerz, M., Schimmelmann, A., 2012. Characterization of chemical functional groups in macerals across different coal ranks via micro-FTIR spectroscopy. Int. J. Coal Geol. 104, 22–33. Clarkson, C.R., Bustin, R.M., 1996. Variation in micropore capacity and size distribution with composition in bituminous coal of the Western Canadian Sedimentary Basin: implications for coalbed methane potential. Fuel 75, 1483–1498. Fielding, C.R., Kassan, J., Draper, J., 1995. Geology of the Bowen and Surat Basins, Eastern Queensland. Conference Publications for Geological Society of Australia. Flores, R.M., 2014. Coal Composition and Reservoir Characterization-Chapter 5. Gamson, P., Beamish, B., Johnson, D., 1996. Coal microstructure and secondary mineralization: their effect on methane recovery. Geol. Soc. Lond., Spec. Publ. 109, 165–179. Gan, H., Nandi, S., Walker Jr., P., 1972. Nature of the porosity in American coals. Fuel 51, 272–277. Guo, Y., Bustin, R.M., 1998. Micro-FTIR spectroscopy of liptinite macerals in coal. Int. J. Coal Geol. 36, 259–275. Harris, L.A., Yust, C.S., 1976. Transmission electron microscope observations of porosity in coal. Fuel 55, 233–236. Ibarra, J., Juan, R., 1985. Structural changes in humic acids during the coalification process. Fuel 64, 650–656. Ibarra, J., Munoz, E., Moliner, R., 1996. FTIR study of the evolution of coal structure during the coalification process. Org. Geochem. 24, 725–735. ISO 7404-3, 2009. Methods for the petrographic analysis of coals—part 5: Method of determining maceral group composition. In: International Organization for Standardization, Geneva, pp. 7. ISO 7404-5, 2009. Methods for the Petrographic Analysis of Coals—Part 5: Method of Determining Microscopically the Reflectance of Vitrinite. International Organization for Standardization, Geneva, pp. 14. Jing, Z., Balucan, R.D., Underschultz, J.R., Steel, K.M., 2018. Oxidant stimulation for enhancing coal seam permeability: Swelling and solubilisation behaviour of unconfined coal particles in oxidants. Fuel 221, 320–328. Khorasani, G.K., 1987. Oil-prone coals of the Walloon coal measures, Surat Basin, Australia. Geol. Soc. Lond., Spec. Publ. 32, 303–310. Laubach, S., Marrett, R., Olson, J., Scott, A., 1998. Characteristics and origins of coal cleat: a review. Int. J. Coal Geol. 35, 175–207.

5. Conclusion NaClO was determined to be a promising oxidant for coal dissolution, particularly for subbituminous coal used in our experiments when compared to the high volatile A bituminous coal. To examine the change in coal structure after oxidation, tests including swelling and leaching tests, porosity measurement, SEM image analysis and cleat flow cell tests were conducted and the results, when integrated, demonstrate that: 1. Unconfined coal particles from both coal samples show swelling, dissolution and breakage in NaClO. 2. The lower rank Coal S was more reactive with NaClO than Coal B according to loss in coal mass and DOC in solute filtrate. Coal S showed 49% mass loss and 3840 mg/L DOC in the filtrate, while Coal B exhibited 4.5% mass loss and 430 mg/L DOC. 3. NaClO oxidation appears to increase the coal porosity by enlarging the pore size, which is also demonstrated by SEM imaging. 4. NaClO oxidation can increase an artificial channel aperture on the natural coal surface and show the potential to increase the cleat system connectivity, particularly demonstrated by the more reactive Coal S sample. 5. Higher Coal S reactivity with NaClO is proposed to be caused by its lower rank and abundant oxygen content. 43

International Journal of Coal Geology 200 (2018) 36–44

Z. Jing et al.

Okolo, G.N., Neomagus, H.W., Everson, R.C., Roberts, M.J., Bunt, J.R., Sakurovs, R., Mathews, J.P., 2015. Chemical–structural properties of South African bituminous coals: insights from wide angle XRD–carbon fraction analysis, ATR–FTIR, solid state 13 C NMR, and HRTEM techniques. Fuel 158, 779–792. Olsen, T.N., Bratton, T.R., Tanner, K.V., Koepsell, R., 2007. Application of Indirect Fracturing for Efficient Stimulation of Coalbed Methane, Rocky Mountain Oil & Gas Technology Symposium. Society of Petroleum Engineers. Palmer, I., 1992. Review of Coalbed Methane Well Stimulation. In: SPE International Meeting on Petroleum Engineering. Society of Petroleum Engineers (SPE). https:// doi.org/10.2118/22395-MS. Palmer, I., 2010. Coalbed methane completions: a world view. Int. J. Coal Geol. 82, 184–195. Palmer, I.D., Moschovidis, Z.A., Cameron, J.R., 2005. Coal Failure and Consequences for Coalbed Methane Wells. In: SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers. Pickel, W., Kus, J., Flores, D., Kalaitzidis, S., Christanis, K., Cardott, B., Misz-Kennan, M., Rodrigues, S., Hentschel, A., Hamor-Vido, M., 2017. Classification of liptinite–ICCP System 1994. Int. J. Coal Geol. 169, 40–61. Turner, L.G., Steel, K.M., 2016. A study into the effect of cleat demineralisation by hydrochloric acid on the permeability of coal. Journal of Natural Gas Science and Engineering 36, 931–942. Turner, L.G., Steel, K.M., Pell, S., 2013. Novel chemical stimulation techniques to enhance coal permeability for coal seam gas extraction. In: SPE 13URCE: Unconventional Resources Conference and Exhibition-Asia Pacific 2013. Society of Petroleum Engineers (SPE), pp. 1–10. Van Krevelen, D., 1993. Coal: Typology–Physics–Chemistry–Constitution (Coal Science & Technology). Elsevier Science, Amsterdam. Van Niekerk, D., Pugmire, R.J., Solum, M.S., Painter, P.C., Mathews, J.P., 2008. Structural characterization of vitrinite-rich and inertinite-rich Permian-aged South African bituminous coals. Int. J. Coal Geol. 76, 290–300. Van Niekerk, D., Halleck, P.M., Mathews, J.P., 2010. Solvent swelling behavior of Permian-aged South African vitrinite-rich and inertinite-rich coals. Fuel 89, 19–25. Wang, S.-H., Griffiths, P.R., 1985. Resolution enhancement of diffuse reflectance ir spectra of coals by Fourier self-deconvolution: 1. CH stretching and bending modes. Fuel 64, 229–236. Wang, Y.-G., Wei, X.-Y., Yan, H.-L., Liu, F.-J., Li, P., Zong, Z.-M., 2014. Sequential oxidation of Jincheng no. 15 anthracite with aqueous sodium hypochlorite. Fuel Process. Technol. 125, 182–189. Yao, Z.-S., Wei, X.-Y., Lv, J., Liu, F.-J., Huang, Y.-G., Xu, J.-J., Chen, F.-J., Huang, Y., Li, Y., Lu, Y., 2010. Oxidation of Shenfu coal with RuO4 and NaOCl. Energy Fuel 24, 1801–1808. Yu, J., Jiang, Y., Tahmasebi, A., Han, Y., Li, X., Lucas, J., Wall, T., 2014. Coal oxidation under mild conditions: current status and applications. Chem. Eng. Technol. 37, 1635–1644.

Levine, J.R., 1996. Model study of the influence of matrix shrinkage on absolute permeability of coal bed reservoirs. Geol. Soc. Lond., Spec. Publ. 109, 197–212. Liu, S., Harpalani, S., 2013. Permeability prediction of coalbed methane reservoirs during primary depletion. Int. J. Coal Geol. 113, 1–10. Liu, F.-J., Wei, X.-Y., Zhu, Y., Gui, J., Wang, Y.-G., Fan, X., Zhao, Y.-P., Zong, Z.-M., Zhao, W., 2013a. Investigation on structural features of Shengli lignite through oxidation under mild conditions. Fuel 109, 316–324. Liu, F.-J., Wei, X.-Y., Zhu, Y., Wang, Y.-G., Li, P., Fan, X., Zhao, Y.-P., Zong, Z.-M., Zhao, W., Wei, Y.-B., 2013b. Oxidation of Shengli lignite with aqueous sodium hypochlorite promoted by pretreatment with aqueous hydrogen peroxide. Fuel 111, 211–215. Liu, F.-J., Wei, X.-Y., Fan, M., Zong, Z.-M., 2016. Separation and structural characterization of the value-added chemicals from mild degradation of lignites: a review. Appl. Energy 170, 415–436. Lv, J.-H., Wei, X.-Y., Qing, Y., Wang, Y.-H., Wen, Z., Zhu, Y., Wang, Y.-G., Zong, Z.-M., 2014. Insight into the structural features of macromolecular aromatic species in Huolinguole lignite through ruthenium ion-catalyzed oxidation. Fuel 128, 231–239. Magill, D.P., Ramurthy, M., Jordan, R., Nguyen, P.D., 2010. Controlling Coal-Fines Production in Massively Cavitated Openhole Coalbed-Methane Wells, SPE Asia Pacific Oil and Gas Conference and Exhibition. Society of Petroleum Engineers. Mahoney, S.A., Rufford, T.E., Rudolph, V., Liu, K.-Y., Rodrigues, S., Steel, K.M., 2015. Creation of microchannels in Bowen Basin coals using UV laser and reactive ion etching. Int. J. Coal Geol. 144, 48–57. Mahoney, S.A., Rufford, T.E., Johnson, D., Dmyterko, A.S., Rodrigues, S., Esterle, J., Rudolph, V., Steel, K.M., 2017. The effect of rank, lithotype and roughness on contact angle measurements in coal cleats. Int. J. Coal Geol. 179, 302–315. Mastalerz, M., Bustin, R., 1993a. Variation in maceral chemistry within and between coals of varying rank: an electron microprobe and micro-Fourier transform infra-red investigation. J. Microsc. 171, 153–166. Mastalerz, M., Bustin, R.M., 1993b. Electron microprobe and micro-FTIR analyses applied to maceral chemistry. Int. J. Coal Geol. 24, 333–345. Mavor, M., Robinson, J., 1993. Analysis of Coal Gas Reservoir Interference and Cavity Well Tests. Low Permeability Reservoirs Symposium. Society of Petroleum Engineers. Mayo, F.R., 1975. Application of sodium hypochlorite oxidations to the structure of coal. Fuel 54, 273–275. Mazumder, S., Wolf, K.-H., Elewaut, K., Ephraim, R., 2006. Application of X-ray computed tomography for analyzing cleat spacing and cleat aperture in coal samples. Int. J. Coal Geol. 68, 205–222. Miura, K., Mae, K., Okutsu, H., Mizutani, N.-a., 1996. New oxidative degradation method for producing fatty acids in high yields and high selectivity from low-rank coals. Energy Fuel 10, 1196–1201. Moore, T.A., 2012. Coalbed methane: a review. Int. J. Coal Geol. 101, 36–81. Murata, S., Tani, Y., Hiro, M., Kidena, K., Artok, L., Nomura, M., Miyake, M., 2001. Structural analysis of coal through RICO reaction: detailed analysis of heavy fractions. Fuel 80, 2099–2109.

44