Simulation of biogenic coalbed gas from anthracite in the south of Qinshui Basin, China

High-rank coal, such as anthracite, has been considered difficult to generate biogas because of the high coalification degree. Selecting anthracite from Sihe Coal Mine, Qinshui Basin, China, as substrate, this study carried out a simulation experiment of biogas generation for 80 days, the purpose of which was to verify whether anthracite could be biodegraded to produce biogas under laboratory conditions. The results showed that the selected anthracite can be utilized by methanogenic bacteria to produce biogas and the approximate production field was 1.79 mL/g, which was less than that of the lower rank coal of other published studies. The generation process can be divided into a rapid growth stage (0–30 days) and a slow descent stage (30–80 days). CO2 and CH4 are the main components of biogas, although some heavy hydrocarbons were also tested. The CO2 concentrations were low (< 30%), and the δ13C-CH4 values were positive (−39.9‰ to −45.8‰), which suggested that the main biogas generation pathway was acetic fermentation. But at the same time, the concentrations of CH4 and CO2 were mutually increasing and decreasing with the passage of experiment time, and δ13C-CH4 tends to be lighten in the later stage(40–80 days), suggesting that parts of biogenic CH4 were generated by way of CO2 reduction.

basins all over the world (Warwick et al. 2008;Strąpoć et al. 2011;Yun et al. 2012;Yoon et al. 2016). Microorganisms can degrade coal organic matter and convert it into coalbed gas through their own metabolic activities when coal reservoirs have a suitable external environment (Green et al. 2008;Jones et al. 2010;Penner et al. 2010;Beckmann et al. 2011;Haider et al. 2013;Wang and Shao 2019). In the actual production of coalbed gas, many companies (Luca Technologies, Ciris Energy, and NextFuel) in the USA have achieved significant economic benefits by strengthening microbial activity and other ways to increase the production of biogenic coalbed gas (Daniel et al. 2015). Apex Australia injected methanogeic bacteria into coal seams in the Sydney Basin to stimulate biogas generation (Faiz et al. 2003).
The generation of biogenic coalbed gas is a multi-step process involving the common metabolism of bacteria and methanogens (methanogens). Organic compounds in coal are considered the substrates. At the initial stage of bio-transformation, complex organic compounds in coal are decomposed into simpler molecules by hydrolyzed bacteria, such as polyromantic hydrocarbons (PAHs), monoaromatic carboxylic acids, ketones, long-chain alkanes, and long-chain fatty acids. These molecules are then converted to simple compounds such as acetate, CO 2 and H 2 , methanol, and formic acid. There are two main pathways to produce CH 4 gas, namely, CO 2 reduction and acetic acid fermentation (Strąpoć et al. 2011;Park and Liang 2016). Recently, it has been found that there is also an archaea that can directly use benthoxy in coal to produce methane (Mayumi et al. 2016). This kind of archaea is named methylotrophic methanogens, and they can produce methane by anaerobic fermentation with formic acid, methanol, and methylamine as substrates (Zhou et al. 2021).
Lower rank coal is considered to have better biogas generation potential than high-rank coal (Rice and Claypool 1981;Orem and Finkelman 2003). The reason is explained as the weak degree of coal mineralization leaves more side chains in the coal, and these simple structures are easy to be used by methanogens to produce coalbed gas, and hydrogen which is rich in chains is important in enhancing biogenic coalbed gas generation even in the high maturity stage (Strąpoć et al. 2011). In recent years, there have been reports about biogenic coalbed gas generation from peat, lignite, sub-bituminous, and bituminous (Ulrich and Bower 2008;Opara et al. 2012;Haider et al. 2013;Haider et al. 2014;Bao et al. 2016;Robbins et al. 2016;Shao et al. 2018;Wang and Shao 2019). Because of the high degree of coalescence and closer chemical structure, high-rank coal, such as anthracite, was considered as an imperfect substrate of biogenic coalbed gas. However, some scholars still believed that high-rank coal could also be anoxic degraded to produce biogas under appropriate conditions, and they had successfully carried out experiments researches on biogenic coalbed gas from higher rank coal (Fallgren et al. 2013;Wei et al. 2014;Susilawati et al. 2015). Bao et al. (2016) summarized 83 data on biogas production in coal seams and found that biogenic coalbed gas can be generated from various coals with different coal ranks, and sub-bituminous and high volatile C-Bituminous coals have higher biogenic coalbed gas yield than peat and anthracite.
Qinshui Basin is one of the main development bases of coalbed gas industry in China. However, due to the influence of geological factors, the production of coalbed gas is low in many areas (Chen et al. 2018). The high coal rank reservoirs in the south of Qinshui Basin are characterized by poor physical properties and low permeability (Qin et al. 2020;Hu et al. 2019). The generation of biogenic coalbed gas has the significance of reforming the pore structure of coal seam, increasing the permeability of coal seam, and facilitating the desorption and migration of coalbed gas (Xia et al. 2014). Therefore, if the microbial enhanced coalbed gas technology can be successfully applied to high-rank coal, it will be of great significance to increase coalbed gas production, improve coal reservoir permeability, and clean utilization of coal.
Using anthracite from Sihe Coal Mine in the south of Qinshui Basin, China, as substrate, this paper carried out the simulated experiment of biogenic coalbed gas generation and focuses on the geochemistry characteristics in the process. The results contribute to the further understanding of the generation mechanism and provide theoretical guidance for biogasification and clean utilization of higher rank coal.

Anthracite sample collection and the basic properties
Four samples, No. SH1, SH2, SH3, and SH4, were collected from Sihe Coal Mine, Jincheng City, Shanxi Province, China (Fig. 1). The samples collected under the mine were immediately wrapped with thin film and sealed with bags to prevent oxidation. The samples were crushed uniformly, and the pure coal samples were selected manually.
A Zeiss Imager Mim microscope was used to measure maximum vitrinite reflectance (R o,max ). The proximate analysis and elemental analysis were conducted following STM Standard D3173-11, D3175-11, D3174-11, and D5373-08. The maceral content was determined by white light photometer and Zeiss Imager Mim microscope. Five hundred points were recorded for each maceral. These results were shown in Table 1.
The R o,max values are all more than 2.28%, and the average value is 2.56%, indicating that the coal have a high degree of metamorphism and belong to anthracite. The fixed carbon contents in coal samples are relatively high with an average of 82.78%. The anthracite samples are of low-volatile (7.16% to 9.97%) and low-ash (7.62% to 12.95%) content according to the Chinese coal industry standard MT/T849-2000 and GB 15224.1-2004. The sulfur contents are very low and belong to super low sulfur coal. Among the organic macerals in coal, the vitrinite group has the highest content (74.2% to 81.1%), followed by the inerter group (16.1% to 20.2%), and the exinite has not been detected. The average mineral content is 3.8%.

Process and method of simulated experiment of biogenic coalbed gas
The simulation experiment was carried out according to the experimental process of bacterial selection, medium configuration, bacterial enrichment culture, inoculation and grouping experiment, and biogas production, as what is shown in Fig. 2.

Natural gas release of coal samples
Anthracite samples were subjected to a natural gas release process before the simulated experiment. They are initially crushed and then placed in a flask connected to a beaker through a glass tube. Place the flask in a thermostatic water bath (60 °C for 7 days). Bubbles in the water indicate that gas has been resolved, and when no bubbles in the water, the natural analysis process is considered complete.

Bacterial samples and the medium compositions
The source of the bacteria used in the simulation experiment was supplied by the Chinese Academy of Science and was enriched and purified in the mine water of Dananhu Coal Mine of Hami Coalfield, Xinjiang Province, China. The bacteria were placed in a low-temperature anaerobic biochemical incubator and stored at a constant temperature of 7 °C.  R o,max is the maximum vitrinite reflectance ; M ad is the moisture on air dry basis; A d is the ash on dry basis; V daf is the volatile matter on dry and ash-free basis; and FC d is fixed carbon content on dry basis. Elements were determined on a dry and ash-free basis, except for sulfur, which was on a dry basis. V is vitrinite, and I is intrinite in coal maceral composition The medium consisted of K 2 HPO 4 0.4 g, MgCl 2 2.0 g, KH 2 PO 4 0.4 g, yeast extract 1.0 g, NH 4 Cl 1.0 g, resazurin 0.01 g, sodium acetate 2.0 g, KCl 0.2 g, NaCl 2.0 g, trace element solution 10.0 mL, and deionized water 1000 mL. The trace element solution was prepared following the reference of Shao et al. (2018). 1.0 mol/L HCl and 1.0 mol/L NaOH were used to adjust the pH value of the medium to 7.0-7.2. The prepared medium was boiled for 5-10 min to remove the dissolved oxygen in the solution and sterilized at 120 °C for 20 min and then placed it in an anaerobic glove box immediately. Before inoculation, 20 mL of 1% Na 2 S and 5% NaHCO 3 mixed solution and 0.5 g cysteine were injected with sterilized deoxygenation syringe to completely remove the dissolved oxygen in the liquid (Shao et al. 2018).

Bacterial enrichment cultivation
The bacterial source was placed in HZQ-F160 incubator with constant temperature oscillation and incubated at 37 °C for 10 h. The whole process of inoculation and cultivation was held in SYQX-II anaerobic incubator. N 2 was used to replace the air in the anaerobic incubator. Three hundred eighty mL of the chilled medium solution was added to the anaerobic tank, and then 3-5 mL of deoxidizer was added, shaken evenly, and then 6 mL of 2.5 mol/L sodium acetate solution was added, and finally, the cultured bacteria source was added. The inoculated anaerobic flask was placed in the HZQ-F160 constant temperature oscillating incubator for enrichment culture. The temperature was set as constant temperature 37 °C, the rotation speed was 50 R/min, and the culture time was 30 days.

Inoculation and grouping experiment
A 500-mL brine bottle with a butyl rubber stopper and wax seal was selected to serve as the simulation experiment device to ensure good airtightness. The assembly of the culture gas collection device is shown in Fig. 3. It was composed of a culture bottle, a needle, a 3-way valve, a 2-way valve, and a 10-mL screw syringe. The gas was collected by the downward drainage saturated brine method.
The experiment was divided into 10 groups, including two blank groups (BK-1, BK-2) and 8 experimental groups (all coal samples were subjected to parallel experiment, SH1-1, SH1-2, SH2-1, SH2-2…). In the 8 experimental groups, 40 g of anthracite, 40 mL of bacterial seed solution, and 350 mL of culture medium were added, respectively. No anthracite was added to the two blank groups. All the culture bottles were placed in HZQ-F160 constant temperature oscillation incubator. The temperature was set at 37 °C, the rotation speed was 50 R/min, and the expected culture time was 80 days. The generated gases were collected on periodically during the experiment (10 days, 20 days, 30 days, 40 days, 60 days, 80 days). The test values below, including gas production, gas composition, and stable isotope composition, are the average values of the corresponding two parallel samples.

Gas analysis
The gas composition was directly determined by East West Electronic Chromatograph (model GC-4000A), and the analysis process has been described in a published paper (Shao et al. 2018).
Stable isotope composition of C (δ 13 C) was tested following the Chinese standard GB/T18340.2-2010 with mass spectrometer (model DELAT V) and HP-PLOT chromatographic column. The split ratio was 15:1. The injection temperature was 40°C. δ 13 C = 1000 (R 1 /R 2 − 1), where R 1 and R 2 are the 13 C/ 12 C ratios of the samples and standard, respectively.

Experimental results
The simulated gas productions in different days are listed in Table 2. According to the previous research results, the culture medium can generate biogas (Wang et al. 2017;Shao et al. 2018). The gas production results of BK group in Table 2 also verify this. Therefore, the biogas production of anthracite was obtained by deducting the gas production from culture medium, as shown in the Eq. (1).
The gas compositions are shown in Table 3. The collected gas mainly consists of N 2 , CO 2 , and CH 4 but also contains little heavy hydrocarbon, including C 2 H 6 and C 3 H 8 . Previous studies have shown that N 2 in the collected gas does not belong to the part of biogas generation (Tao et al. 2007;Shao et al. 2018). The existence of N 2 is caused by the air entering when replacing the air in the process of inoculation. The actual biogas component can be obtained after N 2 is deducted by Eq. (2): Notes: q net is net biogenic gas generated from anthracite samples, mL; q(E) is q of experimental groups, mL; and (1) q(BK) is q of blank group, mL; q is production of biogas, mL; q t is production of collected gas, mL; and C(N 2 ) is concentration of N 2 in Table 3, %. The calculated net biogas productions (q net ) of anthracite are shown in Table 4.

Anthracite biogas production
According to the weight of coal and gas collection time, the net gas yield was determined by Eq. (3): Notes: R net is net biogas yield, mL·g −1 ·day −1 ; m is weight of anthracite samples, g; and Δt is interval time of gas collection.
On day 60, the BK no longer produced biogas, indicating that microorganisms had used the culture solution completely then (Fig. 4a). After 60 days, biogas was still produced in the experimental group, indicating that the anthracite samples in Sihe Mine could be effectively utilized by methanogenic bacteria to produce biogas. Anthracite is characterized by high degree of coalescence and tighter chemical structure, so it is considered by most scholars that it is difficult to biodegrade to produce gas. However, some scholars have successfully simulated gas production from anthracite coal in the laboratory (Xiao et al. 2013). The results of this study once again verify this conclusion.
The total productions of net biogas generated from SH1, SH2, SH3, and SH4 were 65.42, 73.07, 87.52, and 71.63 mL, and the average value was 74.34 mL. This means that approximately 1.79 mL/g of biogas can be produced from anthracite in Sihe Mine, Qinshui Basin. Under similar cultured condition, bituminous coals can produce 5.98 mL/g approximately, and lignite can produce 8.4 mL/g (Shao et al. 2018). Anthracite produce less biogas than that of lower rank coal. There is no doubt that low-rank coal can be easily biodegraded than high-rank coal, because they have more aliphatic hydrocarbons and heteromatons. By the way, the maceral composition was considered to influence the biogas generation rate, which could explain why high-rank coal has (3) R net = q net ∕(m × Δt) more biogas rate than low-rank coal (Haider et al. 2013;Shao et al. 2018).
The whole biogas generation process can be divided into 2 stages: a fast growth stage (0-30 days) and a slow decline stage (30-80 days) (as what is shown in Fig. 4b and c). Similar periodic characteristics occur in almost all biogenic coal bed gas generation simulation experiments Wang and Shao 2019;Shao et al. 2019). The results in this study indicated that the biodegradation process of anthracite is similar to that of low-rank coal. In the first 30 days, the nutrients were rich in organic carbon, and many small soluble organic compounds were available for biodegradation. These organisms effectively activated methanogens in this first stage. The slow decline stage resulted from the low activity of methanogens, which perhaps be due to the inhibitory compounds accumulated in the solution, such as volatile fatty acids or heavy metals (Strąpoć et al. 2011;Wawrik et al. 2012;Fallgren et al. 2013;Chen et al. 2017Chen et al. , 2018. The amount of these substances was not sufficient to completely inhibit methanogenic activity. With the increasing activity of methanogens, they were decomposed to produce biogas.

Simulated biogas composition characteristics
As shown in Table 3, the main components of biogas were CO 2 and CH 4 ; heavy hydrocabons were little. Heavy hydrocarbons were rarely found in biogenic gas simulated experiments of lignite or bituminous coals (Shao et al. 2018. However, they were often found in secondary biogenic coalbed gas reservoirs (Scott et al. 1994;Faiz et al. 2007;Lan et al. 2012). The parent material of gas generation is considered to be directly related to the origin of heavy hydrocarbon, in the early report (Qin et al. 1998). Some study considered that the high content of hydrogen-rich vitrinite and higher coal rank might be the reasons for the high content of heavy hydrocarbons . According to the latest report, heavy hydrocarbons might also be caused by the catalysis of heavy metals (Wu 2011;Wang et al. 2017). However, we considered that the high coalification degree and large molecular structure of anthracite were the reasons for the formation of heavy hydrocarbons. These are all subject to further study.
The change of CO 2 concentrations can be divided into two stages: rising (0-40 days) and falling (40-80 days)  (Fig. 5a). In the first 40 days, the increase curve of CO 2 concentration was oscillatory. On the 30th day, the concentration of CO 2 showed the lowest value, but during the next 10 days, it rapidly increased to the maximum. In the falling stage, CO 2 content decreased almost linearly. CH 4 concentration increased slowly in the first 20 days but increased rapidly in the 20-30 days. During 30-80 days, it increased steadily (Fig. 5b). Concentrations of CO 2 and CH 4 showed an opposite alternation, suggesting that some CO 2 was used to produce CH 4 especially in 30-80 days. This means that part of biogenic CH 4 from anthracite was generated by way of CO 2 reduction in this study. The main pathway was acetic fermentation, because the contents of CO 2 were all lower than 30%, but the CH 4 content was more than 60% in the final 50days.

Carbon isotopic composition of simulated biogas
δ 13 C is an important geochemical parameter of biogas, which can reflect the source and formation mechanism of biogas generation (Hayes 1993;Conrad et al. 2011). Carbon isotope fractionation factor (α C ) is also regarded as identifiers of methanogenic pathway (Hayes 1993). It can be calculated using Eq. (4). Isotope effect (ε C ), which is the difference between C isotope composition CO 2 (δ 13 C-CO2 ) and CH 4 (δ 13 C-CH4 ), was calculated following Eq. (5). It can reflect whether the biogas is produced by acetate fermentation or CO 2 reduction (Hayes 1993). They were all shown in Table 5: These four anthracite samples have similar carbon isotope compositions with the average value of −22.3‰. The biogenic gas has the characteristic of light δ 13 C -CH4 , which was generally considered to range from −110 to −50‰. However, research results in recent years showed that the δ 13 C -CH4 values of coal bed biogas generated under laboratory conditions were higher than −50‰ (Krüger et al. 2008;Shao et al. 2018Shao et al. , 2019. Moreover, acetoclastic methanogenesis can produce 13 C-enriched CH 4 , and the δ 13 C -CH4 values can reach −40‰ (Penning et al. 2006;Vavilin et al. 2008). In this study, the δ 13 C -CH4 values of gas production from anthracite ranged from −39.9 to −45.8‰, which can be considered as biogenic gas and mainly produced by pathway of acetic fermentation. On the 40th day, the mean δ 13 C -CH4 (4) C = ( 13 C −CO 2 + 1000)∕( 13 C −CH 4 + 1000) value was −43.6‰, which was lighter than that of 20 days (−40‰). So pathway of CO 2 reduction was involved at the later stage of the simulation. Figure 6 shows the changes of the mean values of isotopic fractionation parameters of the four samples in the experimental process. δ 13 C -CO2 and α C did not change much, but δ 13 C -CH4 became lighter at the later stage of the   experiment (40 days). Previous simulation results showed that the δ 13 C 1 value of coal seam biogas tends to get lighter with the extension of simulation time, which is consistent with the results of this study. One reason of lightening of δ 13 C 1 can be explained by the CH 4 generation pathway. The range of δ 13 C 1 values of CH 4 generated by different production pathways is different, and the δ 13 C 1 values generated by CO 2 reduction are lower than that of acetic fermentation (Whiticar 1999). The gas compositions and the C isotope compositions can both indicate the generation pathway of biogas, and they also had a unified conclusion in this study. In the early stage (0-40 days), the CO 2 content was high, and CH 4 was mainly generated by acetic acid fermentation at this time, which also results in the heavier δ 13 C 1 value (−40.2‰) at this time. At the late stage of simulation (40-80 days), the joining in of CO 2 reduction pathway made CO 2 concentration decrease and δ 13 C 1 value became lighter (−43.18‰).
However, the formation of biogenic CH 4 in this study was mainly through acetic acid fermentation. α C and ε C values can indicate the results above more accurately. The four experimental groups all produced α C = 1.02. The mean ε C values at 20 days and 40 days were 20.25 and 24.12, respectively. All these values were consistent with what for acetic fermentation pathway (Penning et al. 2006;Goevert and Conrad 2009).

Conclusions
Four anthracite samples from Sihe Mining, in Qinshui Basin, China, were used as the substrates to carry out simulation experiments of biogas generation under laboratory conditions for 80 days. Biogas production, gas composition, and C stable isotope composition were tested to analyze the generation process and pathway of biogas from anthracite. The main conclusions are as follows.
(1) The methanogenic bacteria could use anthracite as substrate to produce biogas. Approximately 1.79 mL/g of biogas were generated from anthracite. Although this value was lower than that of lower rank coal, it is of great significance for the clean utilization of anthracite. The generation process could be divided into two stages: rapid growth stage (0-30 days) and stable decline stage (30-80 days).
(2) CO 2 and CH 4 were the main components of the simulated biogas. The heavy hydrocarbon content is weak. CO 2 content increased rapidly in the early stage (0-40 days) of simulation experiment but decreased continuously in the later stage (40-80 days). The change of CH 4 content was opposite to that of CO 2 .
(3) The δ 13 C -CH4 values of gas production from anthracite ranged from −39.9 to −45.8‰, and δ 13 C -CO2 values ranged from −20.9 to −18.6‰. δ 13 C -CO2 values did not change much, but δ 13 C -CH4 values became lighter (−43.18‰) at the later stage (40 days) than that (−40.2‰) in early stage (20 days) of the experiment. (4) Both of the gas compositions and C isotope compositions verified that the biogenic CH 4 was mainly produced by pathway of acetic acid fermentation, although the pathway of CO 2 reduction joined in the later stage.